From 816626664ccb08368e293a51316f1da815813746 Mon Sep 17 00:00:00 2001 From: Ahmed Kallala Date: Wed, 17 Jun 2026 03:01:19 +0200 Subject: [PATCH] Ahmed moreilp --- measurements/memory/arm/1.log | 37 ++++++++++++ measurements/memory/arm/10.log | 37 ++++++++++++ measurements/memory/arm/100.log | 37 ++++++++++++ measurements/memory/arm/11.log | 37 ++++++++++++ measurements/memory/arm/12.log | 37 ++++++++++++ measurements/memory/arm/13.log | 37 ++++++++++++ measurements/memory/arm/14.log | 37 ++++++++++++ measurements/memory/arm/15.log | 37 ++++++++++++ measurements/memory/arm/16.log | 37 ++++++++++++ measurements/memory/arm/17.log | 37 ++++++++++++ measurements/memory/arm/18.log | 37 ++++++++++++ measurements/memory/arm/19.log | 37 ++++++++++++ measurements/memory/arm/2.log | 37 ++++++++++++ measurements/memory/arm/20.log | 37 ++++++++++++ measurements/memory/arm/21.log | 37 ++++++++++++ measurements/memory/arm/22.log | 37 ++++++++++++ measurements/memory/arm/23.log | 37 ++++++++++++ measurements/memory/arm/24.log | 37 ++++++++++++ measurements/memory/arm/25.log | 37 ++++++++++++ measurements/memory/arm/26.log | 37 ++++++++++++ measurements/memory/arm/27.log | 37 ++++++++++++ measurements/memory/arm/28.log | 37 ++++++++++++ measurements/memory/arm/29.log | 37 ++++++++++++ measurements/memory/arm/3.log | 37 ++++++++++++ measurements/memory/arm/30.log | 37 ++++++++++++ measurements/memory/arm/31.log | 37 ++++++++++++ measurements/memory/arm/32.log | 37 ++++++++++++ measurements/memory/arm/33.log | 37 ++++++++++++ measurements/memory/arm/34.log | 37 ++++++++++++ measurements/memory/arm/35.log | 37 ++++++++++++ measurements/memory/arm/36.log | 37 ++++++++++++ measurements/memory/arm/37.log | 37 ++++++++++++ measurements/memory/arm/38.log | 37 ++++++++++++ measurements/memory/arm/39.log | 37 ++++++++++++ measurements/memory/arm/4.log | 37 ++++++++++++ measurements/memory/arm/40.log | 37 ++++++++++++ measurements/memory/arm/41.log | 37 ++++++++++++ measurements/memory/arm/42.log | 37 ++++++++++++ measurements/memory/arm/43.log | 37 ++++++++++++ measurements/memory/arm/44.log | 37 ++++++++++++ measurements/memory/arm/45.log | 37 ++++++++++++ measurements/memory/arm/46.log | 37 ++++++++++++ measurements/memory/arm/47.log | 37 ++++++++++++ measurements/memory/arm/48.log | 37 ++++++++++++ measurements/memory/arm/49.log | 37 ++++++++++++ measurements/memory/arm/5.log | 37 ++++++++++++ measurements/memory/arm/50.log | 37 ++++++++++++ measurements/memory/arm/51.log | 37 ++++++++++++ measurements/memory/arm/52.log | 37 ++++++++++++ measurements/memory/arm/53.log | 37 ++++++++++++ measurements/memory/arm/54.log | 37 ++++++++++++ measurements/memory/arm/55.log | 37 ++++++++++++ measurements/memory/arm/56.log | 37 ++++++++++++ measurements/memory/arm/57.log | 37 ++++++++++++ measurements/memory/arm/58.log | 37 ++++++++++++ measurements/memory/arm/59.log | 37 ++++++++++++ measurements/memory/arm/6.log | 37 ++++++++++++ measurements/memory/arm/60.log | 37 ++++++++++++ measurements/memory/arm/61.log | 37 ++++++++++++ measurements/memory/arm/62.log | 37 ++++++++++++ measurements/memory/arm/63.log | 37 ++++++++++++ measurements/memory/arm/64.log | 37 ++++++++++++ measurements/memory/arm/65.log | 37 ++++++++++++ measurements/memory/arm/66.log | 37 ++++++++++++ measurements/memory/arm/67.log | 37 ++++++++++++ measurements/memory/arm/68.log | 37 ++++++++++++ measurements/memory/arm/69.log | 37 ++++++++++++ measurements/memory/arm/7.log | 37 ++++++++++++ measurements/memory/arm/70.log | 37 ++++++++++++ measurements/memory/arm/71.log | 37 ++++++++++++ measurements/memory/arm/72.log | 37 ++++++++++++ measurements/memory/arm/73.log | 37 ++++++++++++ measurements/memory/arm/74.log | 37 ++++++++++++ measurements/memory/arm/75.log | 37 ++++++++++++ measurements/memory/arm/76.log | 37 ++++++++++++ measurements/memory/arm/77.log | 37 ++++++++++++ measurements/memory/arm/78.log | 37 ++++++++++++ measurements/memory/arm/79.log | 37 ++++++++++++ measurements/memory/arm/8.log | 37 ++++++++++++ measurements/memory/arm/80.log | 37 ++++++++++++ measurements/memory/arm/81.log | 37 ++++++++++++ measurements/memory/arm/82.log | 37 ++++++++++++ measurements/memory/arm/83.log | 37 ++++++++++++ measurements/memory/arm/84.log | 37 ++++++++++++ measurements/memory/arm/85.log | 37 ++++++++++++ measurements/memory/arm/86.log | 37 ++++++++++++ measurements/memory/arm/87.log | 37 ++++++++++++ measurements/memory/arm/88.log | 37 ++++++++++++ measurements/memory/arm/89.log | 37 ++++++++++++ measurements/memory/arm/9.log | 37 ++++++++++++ measurements/memory/arm/90.log | 37 ++++++++++++ measurements/memory/arm/91.log | 37 ++++++++++++ measurements/memory/arm/92.log | 37 ++++++++++++ measurements/memory/arm/93.log | 37 ++++++++++++ measurements/memory/arm/94.log | 37 ++++++++++++ measurements/memory/arm/95.log | 37 ++++++++++++ measurements/memory/arm/96.log | 37 ++++++++++++ measurements/memory/arm/97.log | 37 ++++++++++++ measurements/memory/arm/98.log | 37 ++++++++++++ measurements/memory/arm/99.log | 37 ++++++++++++ measurements/variant_baseline/arm/1.log | 30 ++++++---- measurements/variant_baseline/arm/2.log | 58 +++++++++++++++++++ measurements/variant_baseline/arm/3.log | 58 +++++++++++++++++++ measurements/variant_branchless/arm/1.log | 28 ++++----- measurements/variant_branchless/arm/2.log | 58 +++++++++++++++++++ measurements/variant_branchless/arm/3.log | 58 +++++++++++++++++++ .../variant_branchless_sqrtfree/arm/1.log | 28 ++++----- .../variant_branchless_sqrtfree/arm/2.log | 58 +++++++++++++++++++ .../variant_branchless_sqrtfree/arm/3.log | 58 +++++++++++++++++++ measurements/variant_cutoff/arm/1.log | 28 ++++----- measurements/variant_cutoff/arm/2.log | 58 +++++++++++++++++++ measurements/variant_cutoff/arm/3.log | 58 +++++++++++++++++++ .../variant_cutoff_callsite/arm/1.log | 24 ++++---- .../variant_cutoff_callsite/arm/2.log | 58 +++++++++++++++++++ .../variant_cutoff_callsite/arm/3.log | 58 +++++++++++++++++++ measurements/variant_float_baseline/arm/1.log | 28 ++++----- measurements/variant_float_baseline/arm/2.log | 58 +++++++++++++++++++ measurements/variant_float_baseline/arm/3.log | 58 +++++++++++++++++++ measurements/variant_float_inlining/arm/1.log | 24 ++++---- measurements/variant_float_inlining/arm/2.log | 58 +++++++++++++++++++ measurements/variant_float_inlining/arm/3.log | 58 +++++++++++++++++++ .../variant_float_inlining_SoA/arm/1.log | 28 ++++----- .../variant_float_inlining_SoA/arm/2.log | 58 +++++++++++++++++++ .../variant_float_inlining_SoA/arm/3.log | 58 +++++++++++++++++++ .../arm/1.log | 28 ++++----- .../arm/2.log | 58 +++++++++++++++++++ .../arm/3.log | 58 +++++++++++++++++++ .../arm/1.log | 26 ++++----- .../arm/2.log | 58 +++++++++++++++++++ .../arm/3.log | 58 +++++++++++++++++++ .../arm/1.log | 28 ++++----- .../arm/2.log | 58 +++++++++++++++++++ .../arm/3.log | 58 +++++++++++++++++++ .../arm/1.log | 22 +++---- .../arm/2.log | 58 +++++++++++++++++++ .../arm/3.log | 58 +++++++++++++++++++ .../arm/1.log | 26 ++++----- .../arm/2.log | 58 +++++++++++++++++++ .../arm/3.log | 58 +++++++++++++++++++ measurements/variant_inlining_7/arm/1.log | 26 ++++----- measurements/variant_inlining_7/arm/2.log | 58 +++++++++++++++++++ measurements/variant_inlining_7/arm/3.log | 58 +++++++++++++++++++ measurements/variant_inlining_7_SoA/arm/1.log | 20 +++---- measurements/variant_inlining_7_SoA/arm/2.log | 58 +++++++++++++++++++ measurements/variant_inlining_7_SoA/arm/3.log | 58 +++++++++++++++++++ .../variant_inlining_7_symmetry_1/arm/1.log | 26 ++++----- .../variant_inlining_7_symmetry_1/arm/2.log | 58 +++++++++++++++++++ .../variant_inlining_7_symmetry_1/arm/3.log | 58 +++++++++++++++++++ .../variant_inlining_7_symmetry_2/arm/1.log | 22 +++---- .../variant_inlining_7_symmetry_2/arm/2.log | 58 +++++++++++++++++++ .../variant_inlining_7_symmetry_2/arm/3.log | 58 +++++++++++++++++++ .../arm/1.log | 24 ++++---- .../arm/2.log | 58 +++++++++++++++++++ .../arm/3.log | 58 +++++++++++++++++++ .../arm/1.log | 26 ++++----- .../arm/2.log | 58 +++++++++++++++++++ .../arm/3.log | 58 +++++++++++++++++++ measurements/variant_inlining_8/arm/1.log | 26 ++++----- measurements/variant_inlining_8/arm/2.log | 58 +++++++++++++++++++ measurements/variant_inlining_8/arm/3.log | 58 +++++++++++++++++++ measurements/variant_lut_exp/arm/1.log | 26 ++++----- measurements/variant_lut_exp/arm/2.log | 58 +++++++++++++++++++ measurements/variant_lut_exp/arm/3.log | 58 +++++++++++++++++++ ...ant_float_inlining_SoA_symmetry_ILP_16.cpp | 1 + ...iant_float_inlining_SoA_symmetry_ILP_2.cpp | 1 + ...iant_float_inlining_SoA_symmetry_ILP_4.cpp | 1 + ...iant_float_inlining_SoA_symmetry_ILP_8.cpp | 1 + 167 files changed, 6415 insertions(+), 269 deletions(-) create mode 100644 measurements/memory/arm/1.log create mode 100644 measurements/memory/arm/10.log create mode 100644 measurements/memory/arm/100.log create mode 100644 measurements/memory/arm/11.log create mode 100644 measurements/memory/arm/12.log create mode 100644 measurements/memory/arm/13.log create mode 100644 measurements/memory/arm/14.log create mode 100644 measurements/memory/arm/15.log create mode 100644 measurements/memory/arm/16.log create mode 100644 measurements/memory/arm/17.log create mode 100644 measurements/memory/arm/18.log create mode 100644 measurements/memory/arm/19.log create mode 100644 measurements/memory/arm/2.log create mode 100644 measurements/memory/arm/20.log create mode 100644 measurements/memory/arm/21.log create mode 100644 measurements/memory/arm/22.log create mode 100644 measurements/memory/arm/23.log create mode 100644 measurements/memory/arm/24.log create mode 100644 measurements/memory/arm/25.log create mode 100644 measurements/memory/arm/26.log create mode 100644 measurements/memory/arm/27.log create mode 100644 measurements/memory/arm/28.log create mode 100644 measurements/memory/arm/29.log create mode 100644 measurements/memory/arm/3.log create mode 100644 measurements/memory/arm/30.log create mode 100644 measurements/memory/arm/31.log create mode 100644 measurements/memory/arm/32.log create mode 100644 measurements/memory/arm/33.log create mode 100644 measurements/memory/arm/34.log create mode 100644 measurements/memory/arm/35.log create mode 100644 measurements/memory/arm/36.log create mode 100644 measurements/memory/arm/37.log create mode 100644 measurements/memory/arm/38.log create mode 100644 measurements/memory/arm/39.log create mode 100644 measurements/memory/arm/4.log create mode 100644 measurements/memory/arm/40.log create mode 100644 measurements/memory/arm/41.log create mode 100644 measurements/memory/arm/42.log create mode 100644 measurements/memory/arm/43.log create mode 100644 measurements/memory/arm/44.log create mode 100644 measurements/memory/arm/45.log create mode 100644 measurements/memory/arm/46.log create mode 100644 measurements/memory/arm/47.log create mode 100644 measurements/memory/arm/48.log create mode 100644 measurements/memory/arm/49.log create mode 100644 measurements/memory/arm/5.log create mode 100644 measurements/memory/arm/50.log create mode 100644 measurements/memory/arm/51.log create mode 100644 measurements/memory/arm/52.log create mode 100644 measurements/memory/arm/53.log create mode 100644 measurements/memory/arm/54.log create mode 100644 measurements/memory/arm/55.log create mode 100644 measurements/memory/arm/56.log create mode 100644 measurements/memory/arm/57.log create mode 100644 measurements/memory/arm/58.log create mode 100644 measurements/memory/arm/59.log create mode 100644 measurements/memory/arm/6.log create mode 100644 measurements/memory/arm/60.log create mode 100644 measurements/memory/arm/61.log create mode 100644 measurements/memory/arm/62.log create mode 100644 measurements/memory/arm/63.log create mode 100644 measurements/memory/arm/64.log create mode 100644 measurements/memory/arm/65.log create mode 100644 measurements/memory/arm/66.log create mode 100644 measurements/memory/arm/67.log create mode 100644 measurements/memory/arm/68.log create mode 100644 measurements/memory/arm/69.log create mode 100644 measurements/memory/arm/7.log create mode 100644 measurements/memory/arm/70.log create mode 100644 measurements/memory/arm/71.log create mode 100644 measurements/memory/arm/72.log create mode 100644 measurements/memory/arm/73.log create mode 100644 measurements/memory/arm/74.log create mode 100644 measurements/memory/arm/75.log create mode 100644 measurements/memory/arm/76.log create mode 100644 measurements/memory/arm/77.log create mode 100644 measurements/memory/arm/78.log create mode 100644 measurements/memory/arm/79.log create mode 100644 measurements/memory/arm/8.log create mode 100644 measurements/memory/arm/80.log create mode 100644 measurements/memory/arm/81.log create mode 100644 measurements/memory/arm/82.log create mode 100644 measurements/memory/arm/83.log create mode 100644 measurements/memory/arm/84.log create mode 100644 measurements/memory/arm/85.log create mode 100644 measurements/memory/arm/86.log create mode 100644 measurements/memory/arm/87.log create mode 100644 measurements/memory/arm/88.log create mode 100644 measurements/memory/arm/89.log create mode 100644 measurements/memory/arm/9.log create mode 100644 measurements/memory/arm/90.log create mode 100644 measurements/memory/arm/91.log create mode 100644 measurements/memory/arm/92.log create mode 100644 measurements/memory/arm/93.log create mode 100644 measurements/memory/arm/94.log create mode 100644 measurements/memory/arm/95.log create mode 100644 measurements/memory/arm/96.log create mode 100644 measurements/memory/arm/97.log create mode 100644 measurements/memory/arm/98.log create mode 100644 measurements/memory/arm/99.log create mode 100644 measurements/variant_baseline/arm/2.log create mode 100644 measurements/variant_baseline/arm/3.log create mode 100644 measurements/variant_branchless/arm/2.log create mode 100644 measurements/variant_branchless/arm/3.log create mode 100644 measurements/variant_branchless_sqrtfree/arm/2.log create mode 100644 measurements/variant_branchless_sqrtfree/arm/3.log create mode 100644 measurements/variant_cutoff/arm/2.log create mode 100644 measurements/variant_cutoff/arm/3.log create mode 100644 measurements/variant_cutoff_callsite/arm/2.log create mode 100644 measurements/variant_cutoff_callsite/arm/3.log create mode 100644 measurements/variant_float_baseline/arm/2.log create mode 100644 measurements/variant_float_baseline/arm/3.log create mode 100644 measurements/variant_float_inlining/arm/2.log create mode 100644 measurements/variant_float_inlining/arm/3.log create mode 100644 measurements/variant_float_inlining_SoA/arm/2.log create mode 100644 measurements/variant_float_inlining_SoA/arm/3.log create mode 100644 measurements/variant_float_inlining_SoA_symmetry/arm/2.log create mode 100644 measurements/variant_float_inlining_SoA_symmetry/arm/3.log create mode 100644 measurements/variant_float_inlining_SoA_symmetry_ILP_16/arm/2.log create mode 100644 measurements/variant_float_inlining_SoA_symmetry_ILP_16/arm/3.log create mode 100644 measurements/variant_float_inlining_SoA_symmetry_ILP_2/arm/2.log create mode 100644 measurements/variant_float_inlining_SoA_symmetry_ILP_2/arm/3.log create mode 100644 measurements/variant_float_inlining_SoA_symmetry_ILP_4/arm/2.log create mode 100644 measurements/variant_float_inlining_SoA_symmetry_ILP_4/arm/3.log create mode 100644 measurements/variant_float_inlining_SoA_symmetry_ILP_8/arm/2.log create mode 100644 measurements/variant_float_inlining_SoA_symmetry_ILP_8/arm/3.log create mode 100644 measurements/variant_inlining_7/arm/2.log create mode 100644 measurements/variant_inlining_7/arm/3.log create mode 100644 measurements/variant_inlining_7_SoA/arm/2.log create mode 100644 measurements/variant_inlining_7_SoA/arm/3.log create mode 100644 measurements/variant_inlining_7_symmetry_1/arm/2.log create mode 100644 measurements/variant_inlining_7_symmetry_1/arm/3.log create mode 100644 measurements/variant_inlining_7_symmetry_2/arm/2.log create mode 100644 measurements/variant_inlining_7_symmetry_2/arm/3.log create mode 100644 measurements/variant_inlining_7_symmetry_2_ilp_1/arm/2.log create mode 100644 measurements/variant_inlining_7_symmetry_2_ilp_1/arm/3.log create mode 100644 measurements/variant_inlining_7_symmetry_2_ilp_2/arm/2.log create mode 100644 measurements/variant_inlining_7_symmetry_2_ilp_2/arm/3.log create mode 100644 measurements/variant_inlining_8/arm/2.log create mode 100644 measurements/variant_inlining_8/arm/3.log create mode 100644 measurements/variant_lut_exp/arm/2.log create mode 100644 measurements/variant_lut_exp/arm/3.log diff --git a/measurements/memory/arm/1.log b/measurements/memory/arm/1.log new file mode 100644 index 0000000..ce80854 --- /dev/null +++ b/measurements/memory/arm/1.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 21.92 per second) + +11264.00 MiB transferred (22441.24 MiB/sec) + + +General statistics: + total time: 0.5016s + total number of events: 11 + +Latency (ms): + min: 500.79 + avg: 500.91 + max: 501.48 + 95th percentile: 0.00 + sum: 5509.97 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.5009/0.00 + diff --git a/measurements/memory/arm/10.log b/measurements/memory/arm/10.log new file mode 100644 index 0000000..343f624 --- /dev/null +++ b/measurements/memory/arm/10.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 79.11 per second) + +11264.00 MiB transferred (81004.50 MiB/sec) + + +General statistics: + total time: 0.1387s + total number of events: 11 + +Latency (ms): + min: 130.00 + avg: 131.00 + max: 138.65 + 95th percentile: 0.00 + sum: 1441.01 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1310/0.00 + diff --git a/measurements/memory/arm/100.log b/measurements/memory/arm/100.log new file mode 100644 index 0000000..39fabc6 --- /dev/null +++ b/measurements/memory/arm/100.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 58.70 per second) + +11264.00 MiB transferred (60109.29 MiB/sec) + + +General statistics: + total time: 0.1871s + total number of events: 11 + +Latency (ms): + min: 185.04 + avg: 186.12 + max: 187.06 + 95th percentile: 0.00 + sum: 2047.26 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1861/0.00 + diff --git a/measurements/memory/arm/11.log b/measurements/memory/arm/11.log new file mode 100644 index 0000000..d4fddc8 --- /dev/null +++ b/measurements/memory/arm/11.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 41.97 per second) + +11264.00 MiB transferred (42980.72 MiB/sec) + + +General statistics: + total time: 0.2618s + total number of events: 11 + +Latency (ms): + min: 259.41 + avg: 261.39 + max: 261.63 + 95th percentile: 0.00 + sum: 2875.24 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.2614/0.00 + diff --git a/measurements/memory/arm/12.log b/measurements/memory/arm/12.log new file mode 100644 index 0000000..126be80 --- /dev/null +++ b/measurements/memory/arm/12.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 60.97 per second) + +11264.00 MiB transferred (62433.14 MiB/sec) + + +General statistics: + total time: 0.1801s + total number of events: 11 + +Latency (ms): + min: 175.87 + avg: 179.38 + max: 180.04 + 95th percentile: 0.00 + sum: 1973.22 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1794/0.00 + diff --git a/measurements/memory/arm/13.log b/measurements/memory/arm/13.log new file mode 100644 index 0000000..41c4e1e --- /dev/null +++ b/measurements/memory/arm/13.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 73.70 per second) + +11264.00 MiB transferred (75469.68 MiB/sec) + + +General statistics: + total time: 0.1490s + total number of events: 11 + +Latency (ms): + min: 137.47 + avg: 142.05 + max: 148.91 + 95th percentile: 0.00 + sum: 1562.53 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1420/0.00 + diff --git a/measurements/memory/arm/14.log b/measurements/memory/arm/14.log new file mode 100644 index 0000000..da8e024 --- /dev/null +++ b/measurements/memory/arm/14.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 71.13 per second) + +11264.00 MiB transferred (72832.14 MiB/sec) + + +General statistics: + total time: 0.1544s + total number of events: 11 + +Latency (ms): + min: 142.01 + avg: 146.39 + max: 154.20 + 95th percentile: 0.00 + sum: 1610.30 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1464/0.00 + diff --git a/measurements/memory/arm/15.log b/measurements/memory/arm/15.log new file mode 100644 index 0000000..a3ab379 --- /dev/null +++ b/measurements/memory/arm/15.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 61.50 per second) + +11264.00 MiB transferred (62977.33 MiB/sec) + + +General statistics: + total time: 0.1786s + total number of events: 11 + +Latency (ms): + min: 178.10 + avg: 178.19 + max: 178.46 + 95th percentile: 0.00 + sum: 1960.07 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1782/0.00 + diff --git a/measurements/memory/arm/16.log b/measurements/memory/arm/16.log new file mode 100644 index 0000000..4540bb6 --- /dev/null +++ b/measurements/memory/arm/16.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 66.19 per second) + +11264.00 MiB transferred (67776.62 MiB/sec) + + +General statistics: + total time: 0.1659s + total number of events: 11 + +Latency (ms): + min: 156.46 + avg: 159.49 + max: 165.68 + 95th percentile: 0.00 + sum: 1754.44 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1595/0.00 + diff --git a/measurements/memory/arm/17.log b/measurements/memory/arm/17.log new file mode 100644 index 0000000..097a616 --- /dev/null +++ b/measurements/memory/arm/17.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 73.41 per second) + +11264.00 MiB transferred (75172.01 MiB/sec) + + +General statistics: + total time: 0.1496s + total number of events: 11 + +Latency (ms): + min: 144.85 + avg: 145.47 + max: 149.43 + 95th percentile: 0.00 + sum: 1600.12 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1455/0.00 + diff --git a/measurements/memory/arm/18.log b/measurements/memory/arm/18.log new file mode 100644 index 0000000..d2694ff --- /dev/null +++ b/measurements/memory/arm/18.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 76.47 per second) + +11264.00 MiB transferred (78310.32 MiB/sec) + + +General statistics: + total time: 0.1435s + total number of events: 11 + +Latency (ms): + min: 135.95 + avg: 138.94 + max: 143.42 + 95th percentile: 0.00 + sum: 1528.34 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1389/0.00 + diff --git a/measurements/memory/arm/19.log b/measurements/memory/arm/19.log new file mode 100644 index 0000000..b3f997b --- /dev/null +++ b/measurements/memory/arm/19.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 74.49 per second) + +11264.00 MiB transferred (76274.06 MiB/sec) + + +General statistics: + total time: 0.1474s + total number of events: 11 + +Latency (ms): + min: 145.60 + avg: 146.91 + max: 147.28 + 95th percentile: 0.00 + sum: 1616.05 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1469/0.00 + diff --git a/measurements/memory/arm/2.log b/measurements/memory/arm/2.log new file mode 100644 index 0000000..bf35fc2 --- /dev/null +++ b/measurements/memory/arm/2.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 69.94 per second) + +11264.00 MiB transferred (71623.42 MiB/sec) + + +General statistics: + total time: 0.1568s + total number of events: 11 + +Latency (ms): + min: 155.54 + avg: 156.05 + max: 156.66 + 95th percentile: 0.00 + sum: 1716.51 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1560/0.00 + diff --git a/measurements/memory/arm/20.log b/measurements/memory/arm/20.log new file mode 100644 index 0000000..edf4d7d --- /dev/null +++ b/measurements/memory/arm/20.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 84.63 per second) + +11264.00 MiB transferred (86658.15 MiB/sec) + + +General statistics: + total time: 0.1297s + total number of events: 11 + +Latency (ms): + min: 129.45 + avg: 129.52 + max: 129.62 + 95th percentile: 0.00 + sum: 1424.72 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1295/0.00 + diff --git a/measurements/memory/arm/21.log b/measurements/memory/arm/21.log new file mode 100644 index 0000000..7c81da2 --- /dev/null +++ b/measurements/memory/arm/21.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 65.18 per second) + +11264.00 MiB transferred (66747.65 MiB/sec) + + +General statistics: + total time: 0.1685s + total number of events: 11 + +Latency (ms): + min: 167.88 + avg: 168.16 + max: 168.43 + 95th percentile: 0.00 + sum: 1849.71 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1682/0.00 + diff --git a/measurements/memory/arm/22.log b/measurements/memory/arm/22.log new file mode 100644 index 0000000..5ccd692 --- /dev/null +++ b/measurements/memory/arm/22.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 78.55 per second) + +11264.00 MiB transferred (80431.29 MiB/sec) + + +General statistics: + total time: 0.1397s + total number of events: 11 + +Latency (ms): + min: 131.28 + avg: 135.14 + max: 139.68 + 95th percentile: 0.00 + sum: 1486.57 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1351/0.00 + diff --git a/measurements/memory/arm/23.log b/measurements/memory/arm/23.log new file mode 100644 index 0000000..eb23b83 --- /dev/null +++ b/measurements/memory/arm/23.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 70.30 per second) + +11264.00 MiB transferred (71986.86 MiB/sec) + + +General statistics: + total time: 0.1562s + total number of events: 11 + +Latency (ms): + min: 143.48 + avg: 146.66 + max: 156.03 + 95th percentile: 0.00 + sum: 1613.29 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1467/0.00 + diff --git a/measurements/memory/arm/24.log b/measurements/memory/arm/24.log new file mode 100644 index 0000000..d67416f --- /dev/null +++ b/measurements/memory/arm/24.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 71.98 per second) + +11264.00 MiB transferred (73704.74 MiB/sec) + + +General statistics: + total time: 0.1525s + total number of events: 11 + +Latency (ms): + min: 152.23 + avg: 152.30 + max: 152.38 + 95th percentile: 0.00 + sum: 1675.26 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1523/0.00 + diff --git a/measurements/memory/arm/25.log b/measurements/memory/arm/25.log new file mode 100644 index 0000000..b35ecd9 --- /dev/null +++ b/measurements/memory/arm/25.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 64.18 per second) + +11264.00 MiB transferred (65719.15 MiB/sec) + + +General statistics: + total time: 0.1711s + total number of events: 11 + +Latency (ms): + min: 160.31 + avg: 163.34 + max: 171.01 + 95th percentile: 0.00 + sum: 1796.72 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1633/0.00 + diff --git a/measurements/memory/arm/26.log b/measurements/memory/arm/26.log new file mode 100644 index 0000000..b1d4d4f --- /dev/null +++ b/measurements/memory/arm/26.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 73.07 per second) + +11264.00 MiB transferred (74821.48 MiB/sec) + + +General statistics: + total time: 0.1502s + total number of events: 11 + +Latency (ms): + min: 139.86 + avg: 145.30 + max: 150.09 + 95th percentile: 0.00 + sum: 1598.27 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1453/0.00 + diff --git a/measurements/memory/arm/27.log b/measurements/memory/arm/27.log new file mode 100644 index 0000000..1d336a0 --- /dev/null +++ b/measurements/memory/arm/27.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 96.09 per second) + +11264.00 MiB transferred (98391.87 MiB/sec) + + +General statistics: + total time: 0.1142s + total number of events: 11 + +Latency (ms): + min: 113.78 + avg: 113.88 + max: 114.12 + 95th percentile: 0.00 + sum: 1252.70 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1139/0.00 + diff --git a/measurements/memory/arm/28.log b/measurements/memory/arm/28.log new file mode 100644 index 0000000..841d044 --- /dev/null +++ b/measurements/memory/arm/28.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 79.94 per second) + +11264.00 MiB transferred (81859.87 MiB/sec) + + +General statistics: + total time: 0.1372s + total number of events: 11 + +Latency (ms): + min: 131.93 + avg: 133.05 + max: 137.15 + 95th percentile: 0.00 + sum: 1463.57 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1331/0.00 + diff --git a/measurements/memory/arm/29.log b/measurements/memory/arm/29.log new file mode 100644 index 0000000..0403715 --- /dev/null +++ b/measurements/memory/arm/29.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 80.88 per second) + +11264.00 MiB transferred (82818.66 MiB/sec) + + +General statistics: + total time: 0.1357s + total number of events: 11 + +Latency (ms): + min: 119.19 + avg: 125.92 + max: 135.57 + 95th percentile: 0.00 + sum: 1385.13 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1259/0.01 + diff --git a/measurements/memory/arm/3.log b/measurements/memory/arm/3.log new file mode 100644 index 0000000..23fb851 --- /dev/null +++ b/measurements/memory/arm/3.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 71.12 per second) + +11264.00 MiB transferred (72829.31 MiB/sec) + + +General statistics: + total time: 0.1543s + total number of events: 11 + +Latency (ms): + min: 148.58 + avg: 151.23 + max: 154.21 + 95th percentile: 0.00 + sum: 1663.57 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1512/0.00 + diff --git a/measurements/memory/arm/30.log b/measurements/memory/arm/30.log new file mode 100644 index 0000000..4d25adf --- /dev/null +++ b/measurements/memory/arm/30.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 70.38 per second) + +11264.00 MiB transferred (72070.69 MiB/sec) + + +General statistics: + total time: 0.1560s + total number of events: 11 + +Latency (ms): + min: 146.09 + avg: 150.12 + max: 155.15 + 95th percentile: 0.00 + sum: 1651.27 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1501/0.00 + diff --git a/measurements/memory/arm/31.log b/measurements/memory/arm/31.log new file mode 100644 index 0000000..a1203da --- /dev/null +++ b/measurements/memory/arm/31.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 86.10 per second) + +11264.00 MiB transferred (88163.24 MiB/sec) + + +General statistics: + total time: 0.1275s + total number of events: 11 + +Latency (ms): + min: 122.13 + avg: 124.05 + max: 127.35 + 95th percentile: 0.00 + sum: 1364.50 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1240/0.00 + diff --git a/measurements/memory/arm/32.log b/measurements/memory/arm/32.log new file mode 100644 index 0000000..4d242b2 --- /dev/null +++ b/measurements/memory/arm/32.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 82.45 per second) + +11264.00 MiB transferred (84432.72 MiB/sec) + + +General statistics: + total time: 0.1331s + total number of events: 11 + +Latency (ms): + min: 132.82 + avg: 132.89 + max: 132.96 + 95th percentile: 0.00 + sum: 1461.77 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1329/0.00 + diff --git a/measurements/memory/arm/33.log b/measurements/memory/arm/33.log new file mode 100644 index 0000000..284487c --- /dev/null +++ b/measurements/memory/arm/33.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 91.82 per second) + +11264.00 MiB transferred (94024.94 MiB/sec) + + +General statistics: + total time: 0.1195s + total number of events: 11 + +Latency (ms): + min: 113.57 + avg: 116.00 + max: 119.38 + 95th percentile: 0.00 + sum: 1276.05 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1160/0.00 + diff --git a/measurements/memory/arm/34.log b/measurements/memory/arm/34.log new file mode 100644 index 0000000..ef2183e --- /dev/null +++ b/measurements/memory/arm/34.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 62.02 per second) + +11264.00 MiB transferred (63504.95 MiB/sec) + + +General statistics: + total time: 0.1771s + total number of events: 11 + +Latency (ms): + min: 168.02 + avg: 169.00 + max: 176.99 + 95th percentile: 0.00 + sum: 1859.05 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1690/0.00 + diff --git a/measurements/memory/arm/35.log b/measurements/memory/arm/35.log new file mode 100644 index 0000000..4ff9b0a --- /dev/null +++ b/measurements/memory/arm/35.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 86.00 per second) + +11264.00 MiB transferred (88068.12 MiB/sec) + + +General statistics: + total time: 0.1276s + total number of events: 11 + +Latency (ms): + min: 109.70 + avg: 116.80 + max: 127.54 + 95th percentile: 0.00 + sum: 1284.77 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1168/0.01 + diff --git a/measurements/memory/arm/36.log b/measurements/memory/arm/36.log new file mode 100644 index 0000000..d782ee4 --- /dev/null +++ b/measurements/memory/arm/36.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 71.78 per second) + +11264.00 MiB transferred (73503.21 MiB/sec) + + +General statistics: + total time: 0.1529s + total number of events: 11 + +Latency (ms): + min: 149.50 + avg: 150.75 + max: 152.83 + 95th percentile: 0.00 + sum: 1658.21 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1507/0.00 + diff --git a/measurements/memory/arm/37.log b/measurements/memory/arm/37.log new file mode 100644 index 0000000..0ae1602 --- /dev/null +++ b/measurements/memory/arm/37.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 84.09 per second) + +11264.00 MiB transferred (86103.04 MiB/sec) + + +General statistics: + total time: 0.1306s + total number of events: 11 + +Latency (ms): + min: 117.74 + avg: 120.88 + max: 130.47 + 95th percentile: 0.00 + sum: 1329.71 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1209/0.00 + diff --git a/measurements/memory/arm/38.log b/measurements/memory/arm/38.log new file mode 100644 index 0000000..76bf356 --- /dev/null +++ b/measurements/memory/arm/38.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 73.28 per second) + +11264.00 MiB transferred (75039.80 MiB/sec) + + +General statistics: + total time: 0.1498s + total number of events: 11 + +Latency (ms): + min: 138.49 + avg: 142.47 + max: 149.73 + 95th percentile: 0.00 + sum: 1567.16 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1425/0.00 + diff --git a/measurements/memory/arm/39.log b/measurements/memory/arm/39.log new file mode 100644 index 0000000..e3952bc --- /dev/null +++ b/measurements/memory/arm/39.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 68.78 per second) + +11264.00 MiB transferred (70435.66 MiB/sec) + + +General statistics: + total time: 0.1596s + total number of events: 11 + +Latency (ms): + min: 149.51 + avg: 153.34 + max: 159.45 + 95th percentile: 0.00 + sum: 1686.78 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1533/0.00 + diff --git a/measurements/memory/arm/4.log b/measurements/memory/arm/4.log new file mode 100644 index 0000000..ee7d5e7 --- /dev/null +++ b/measurements/memory/arm/4.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 67.61 per second) + +11264.00 MiB transferred (69235.12 MiB/sec) + + +General statistics: + total time: 0.1624s + total number of events: 11 + +Latency (ms): + min: 160.63 + avg: 161.86 + max: 162.26 + 95th percentile: 0.00 + sum: 1780.49 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1619/0.00 + diff --git a/measurements/memory/arm/40.log b/measurements/memory/arm/40.log new file mode 100644 index 0000000..1208677 --- /dev/null +++ b/measurements/memory/arm/40.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 67.34 per second) + +11264.00 MiB transferred (68955.81 MiB/sec) + + +General statistics: + total time: 0.1630s + total number of events: 11 + +Latency (ms): + min: 151.53 + avg: 153.06 + max: 162.85 + 95th percentile: 0.00 + sum: 1683.66 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1531/0.00 + diff --git a/measurements/memory/arm/41.log b/measurements/memory/arm/41.log new file mode 100644 index 0000000..6284c57 --- /dev/null +++ b/measurements/memory/arm/41.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 73.43 per second) + +11264.00 MiB transferred (75192.09 MiB/sec) + + +General statistics: + total time: 0.1495s + total number of events: 11 + +Latency (ms): + min: 141.75 + avg: 144.55 + max: 149.40 + 95th percentile: 0.00 + sum: 1590.09 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1446/0.00 + diff --git a/measurements/memory/arm/42.log b/measurements/memory/arm/42.log new file mode 100644 index 0000000..ec4477c --- /dev/null +++ b/measurements/memory/arm/42.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 94.90 per second) + +11264.00 MiB transferred (97182.20 MiB/sec) + + +General statistics: + total time: 0.1156s + total number of events: 11 + +Latency (ms): + min: 113.18 + avg: 113.43 + max: 115.50 + 95th percentile: 0.00 + sum: 1247.78 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1134/0.00 + diff --git a/measurements/memory/arm/43.log b/measurements/memory/arm/43.log new file mode 100644 index 0000000..beb184c --- /dev/null +++ b/measurements/memory/arm/43.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 63.56 per second) + +11264.00 MiB transferred (65085.75 MiB/sec) + + +General statistics: + total time: 0.1728s + total number of events: 11 + +Latency (ms): + min: 161.91 + avg: 166.26 + max: 172.73 + 95th percentile: 0.00 + sum: 1828.89 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1663/0.00 + diff --git a/measurements/memory/arm/44.log b/measurements/memory/arm/44.log new file mode 100644 index 0000000..2ca5398 --- /dev/null +++ b/measurements/memory/arm/44.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 65.43 per second) + +11264.00 MiB transferred (66995.78 MiB/sec) + + +General statistics: + total time: 0.1679s + total number of events: 11 + +Latency (ms): + min: 164.31 + avg: 165.23 + max: 167.39 + 95th percentile: 0.00 + sum: 1817.56 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1652/0.00 + diff --git a/measurements/memory/arm/45.log b/measurements/memory/arm/45.log new file mode 100644 index 0000000..dacebf8 --- /dev/null +++ b/measurements/memory/arm/45.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 66.39 per second) + +11264.00 MiB transferred (67987.71 MiB/sec) + + +General statistics: + total time: 0.1654s + total number of events: 11 + +Latency (ms): + min: 163.71 + avg: 164.88 + max: 165.29 + 95th percentile: 0.00 + sum: 1813.66 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1649/0.00 + diff --git a/measurements/memory/arm/46.log b/measurements/memory/arm/46.log new file mode 100644 index 0000000..fb84996 --- /dev/null +++ b/measurements/memory/arm/46.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 67.11 per second) + +11264.00 MiB transferred (68721.90 MiB/sec) + + +General statistics: + total time: 0.1636s + total number of events: 11 + +Latency (ms): + min: 153.70 + avg: 156.51 + max: 163.49 + 95th percentile: 0.00 + sum: 1721.62 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1565/0.00 + diff --git a/measurements/memory/arm/47.log b/measurements/memory/arm/47.log new file mode 100644 index 0000000..9f0d2fd --- /dev/null +++ b/measurements/memory/arm/47.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 67.22 per second) + +11264.00 MiB transferred (68831.92 MiB/sec) + + +General statistics: + total time: 0.1634s + total number of events: 11 + +Latency (ms): + min: 155.29 + avg: 159.04 + max: 163.26 + 95th percentile: 0.00 + sum: 1749.45 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1590/0.00 + diff --git a/measurements/memory/arm/48.log b/measurements/memory/arm/48.log new file mode 100644 index 0000000..cfb079b --- /dev/null +++ b/measurements/memory/arm/48.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 67.73 per second) + +11264.00 MiB transferred (69354.48 MiB/sec) + + +General statistics: + total time: 0.1621s + total number of events: 11 + +Latency (ms): + min: 152.80 + avg: 156.15 + max: 161.23 + 95th percentile: 0.00 + sum: 1717.62 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1561/0.00 + diff --git a/measurements/memory/arm/49.log b/measurements/memory/arm/49.log new file mode 100644 index 0000000..673f1d9 --- /dev/null +++ b/measurements/memory/arm/49.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 71.96 per second) + +11264.00 MiB transferred (73685.93 MiB/sec) + + +General statistics: + total time: 0.1526s + total number of events: 11 + +Latency (ms): + min: 143.45 + avg: 146.09 + max: 152.47 + 95th percentile: 0.00 + sum: 1606.96 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1461/0.00 + diff --git a/measurements/memory/arm/5.log b/measurements/memory/arm/5.log new file mode 100644 index 0000000..092de15 --- /dev/null +++ b/measurements/memory/arm/5.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 68.31 per second) + +11264.00 MiB transferred (69944.49 MiB/sec) + + +General statistics: + total time: 0.1608s + total number of events: 11 + +Latency (ms): + min: 159.70 + avg: 160.18 + max: 160.61 + 95th percentile: 0.00 + sum: 1762.00 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1602/0.00 + diff --git a/measurements/memory/arm/50.log b/measurements/memory/arm/50.log new file mode 100644 index 0000000..7d163ba --- /dev/null +++ b/measurements/memory/arm/50.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 58.29 per second) + +11264.00 MiB transferred (59688.84 MiB/sec) + + +General statistics: + total time: 0.1884s + total number of events: 11 + +Latency (ms): + min: 188.06 + avg: 188.13 + max: 188.28 + 95th percentile: 0.00 + sum: 2069.44 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1881/0.00 + diff --git a/measurements/memory/arm/51.log b/measurements/memory/arm/51.log new file mode 100644 index 0000000..8e7c79b --- /dev/null +++ b/measurements/memory/arm/51.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 65.24 per second) + +11264.00 MiB transferred (66801.88 MiB/sec) + + +General statistics: + total time: 0.1683s + total number of events: 11 + +Latency (ms): + min: 167.31 + avg: 168.09 + max: 168.18 + 95th percentile: 0.00 + sum: 1848.95 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1681/0.00 + diff --git a/measurements/memory/arm/52.log b/measurements/memory/arm/52.log new file mode 100644 index 0000000..0f6c138 --- /dev/null +++ b/measurements/memory/arm/52.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 18.35 per second) + +11264.00 MiB transferred (18792.13 MiB/sec) + + +General statistics: + total time: 0.5991s + total number of events: 11 + +Latency (ms): + min: 598.46 + avg: 598.62 + max: 598.75 + 95th percentile: 0.00 + sum: 6584.80 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.5986/0.00 + diff --git a/measurements/memory/arm/53.log b/measurements/memory/arm/53.log new file mode 100644 index 0000000..d562e35 --- /dev/null +++ b/measurements/memory/arm/53.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 68.96 per second) + +11264.00 MiB transferred (70619.36 MiB/sec) + + +General statistics: + total time: 0.1592s + total number of events: 11 + +Latency (ms): + min: 137.05 + avg: 153.30 + max: 159.19 + 95th percentile: 0.00 + sum: 1686.30 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1533/0.01 + diff --git a/measurements/memory/arm/54.log b/measurements/memory/arm/54.log new file mode 100644 index 0000000..fd7aa6a --- /dev/null +++ b/measurements/memory/arm/54.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 79.36 per second) + +11264.00 MiB transferred (81265.15 MiB/sec) + + +General statistics: + total time: 0.1383s + total number of events: 11 + +Latency (ms): + min: 135.10 + avg: 135.81 + max: 137.62 + 95th percentile: 0.00 + sum: 1493.90 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1358/0.00 + diff --git a/measurements/memory/arm/55.log b/measurements/memory/arm/55.log new file mode 100644 index 0000000..60215a7 --- /dev/null +++ b/measurements/memory/arm/55.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 70.89 per second) + +11264.00 MiB transferred (72595.09 MiB/sec) + + +General statistics: + total time: 0.1549s + total number of events: 11 + +Latency (ms): + min: 138.62 + avg: 146.34 + max: 153.72 + 95th percentile: 0.00 + sum: 1609.73 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1463/0.00 + diff --git a/measurements/memory/arm/56.log b/measurements/memory/arm/56.log new file mode 100644 index 0000000..2ee02f3 --- /dev/null +++ b/measurements/memory/arm/56.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 66.24 per second) + +11264.00 MiB transferred (67830.50 MiB/sec) + + +General statistics: + total time: 0.1658s + total number of events: 11 + +Latency (ms): + min: 162.32 + avg: 163.68 + max: 165.70 + 95th percentile: 0.00 + sum: 1800.43 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1637/0.00 + diff --git a/measurements/memory/arm/57.log b/measurements/memory/arm/57.log new file mode 100644 index 0000000..4db6b9a --- /dev/null +++ b/measurements/memory/arm/57.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 80.01 per second) + +11264.00 MiB transferred (81934.30 MiB/sec) + + +General statistics: + total time: 0.1372s + total number of events: 11 + +Latency (ms): + min: 126.74 + avg: 135.96 + max: 137.13 + 95th percentile: 0.00 + sum: 1495.59 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1360/0.00 + diff --git a/measurements/memory/arm/58.log b/measurements/memory/arm/58.log new file mode 100644 index 0000000..0b1a36c --- /dev/null +++ b/measurements/memory/arm/58.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 59.60 per second) + +11264.00 MiB transferred (61031.64 MiB/sec) + + +General statistics: + total time: 0.1843s + total number of events: 11 + +Latency (ms): + min: 178.32 + avg: 182.96 + max: 184.24 + 95th percentile: 0.00 + sum: 2012.53 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1830/0.00 + diff --git a/measurements/memory/arm/59.log b/measurements/memory/arm/59.log new file mode 100644 index 0000000..1ec70f4 --- /dev/null +++ b/measurements/memory/arm/59.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 86.64 per second) + +11264.00 MiB transferred (88718.76 MiB/sec) + + +General statistics: + total time: 0.1266s + total number of events: 11 + +Latency (ms): + min: 124.70 + avg: 125.00 + max: 126.52 + 95th percentile: 0.00 + sum: 1374.95 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1250/0.00 + diff --git a/measurements/memory/arm/6.log b/measurements/memory/arm/6.log new file mode 100644 index 0000000..3f01020 --- /dev/null +++ b/measurements/memory/arm/6.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 87.28 per second) + +11264.00 MiB transferred (89378.38 MiB/sec) + + +General statistics: + total time: 0.1257s + total number of events: 11 + +Latency (ms): + min: 120.82 + avg: 122.18 + max: 125.62 + 95th percentile: 0.00 + sum: 1343.93 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1222/0.00 + diff --git a/measurements/memory/arm/60.log b/measurements/memory/arm/60.log new file mode 100644 index 0000000..ac108c6 --- /dev/null +++ b/measurements/memory/arm/60.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 52.85 per second) + +11264.00 MiB transferred (54118.72 MiB/sec) + + +General statistics: + total time: 0.2078s + total number of events: 11 + +Latency (ms): + min: 192.26 + avg: 204.02 + max: 207.77 + 95th percentile: 0.00 + sum: 2244.18 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.2040/0.01 + diff --git a/measurements/memory/arm/61.log b/measurements/memory/arm/61.log new file mode 100644 index 0000000..0f0fa44 --- /dev/null +++ b/measurements/memory/arm/61.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 60.28 per second) + +11264.00 MiB transferred (61728.33 MiB/sec) + + +General statistics: + total time: 0.1822s + total number of events: 11 + +Latency (ms): + min: 180.99 + avg: 181.89 + max: 182.08 + 95th percentile: 0.00 + sum: 2000.77 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1819/0.00 + diff --git a/measurements/memory/arm/62.log b/measurements/memory/arm/62.log new file mode 100644 index 0000000..4cc487d --- /dev/null +++ b/measurements/memory/arm/62.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 72.66 per second) + +11264.00 MiB transferred (74405.33 MiB/sec) + + +General statistics: + total time: 0.1511s + total number of events: 11 + +Latency (ms): + min: 144.35 + avg: 147.67 + max: 151.02 + 95th percentile: 0.00 + sum: 1624.42 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1477/0.00 + diff --git a/measurements/memory/arm/63.log b/measurements/memory/arm/63.log new file mode 100644 index 0000000..ff55d80 --- /dev/null +++ b/measurements/memory/arm/63.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 79.53 per second) + +11264.00 MiB transferred (81443.78 MiB/sec) + + +General statistics: + total time: 0.1380s + total number of events: 11 + +Latency (ms): + min: 126.94 + avg: 135.74 + max: 137.91 + 95th percentile: 0.00 + sum: 1493.16 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1357/0.00 + diff --git a/measurements/memory/arm/64.log b/measurements/memory/arm/64.log new file mode 100644 index 0000000..6875dcd --- /dev/null +++ b/measurements/memory/arm/64.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 65.48 per second) + +11264.00 MiB transferred (67054.80 MiB/sec) + + +General statistics: + total time: 0.1676s + total number of events: 11 + +Latency (ms): + min: 154.30 + avg: 160.93 + max: 167.51 + 95th percentile: 0.00 + sum: 1770.28 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1609/0.00 + diff --git a/measurements/memory/arm/65.log b/measurements/memory/arm/65.log new file mode 100644 index 0000000..02473ed --- /dev/null +++ b/measurements/memory/arm/65.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 71.45 per second) + +11264.00 MiB transferred (73168.51 MiB/sec) + + +General statistics: + total time: 0.1537s + total number of events: 11 + +Latency (ms): + min: 144.43 + avg: 146.86 + max: 152.90 + 95th percentile: 0.00 + sum: 1615.44 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1469/0.00 + diff --git a/measurements/memory/arm/66.log b/measurements/memory/arm/66.log new file mode 100644 index 0000000..0834f60 --- /dev/null +++ b/measurements/memory/arm/66.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 63.81 per second) + +11264.00 MiB transferred (65341.35 MiB/sec) + + +General statistics: + total time: 0.1721s + total number of events: 11 + +Latency (ms): + min: 159.22 + avg: 160.09 + max: 165.08 + 95th percentile: 0.00 + sum: 1761.02 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1601/0.00 + diff --git a/measurements/memory/arm/67.log b/measurements/memory/arm/67.log new file mode 100644 index 0000000..3eef523 --- /dev/null +++ b/measurements/memory/arm/67.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 23.83 per second) + +11264.00 MiB transferred (24400.44 MiB/sec) + + +General statistics: + total time: 0.4614s + total number of events: 11 + +Latency (ms): + min: 455.80 + avg: 459.69 + max: 461.24 + 95th percentile: 0.00 + sum: 5056.55 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.4597/0.00 + diff --git a/measurements/memory/arm/68.log b/measurements/memory/arm/68.log new file mode 100644 index 0000000..5c13019 --- /dev/null +++ b/measurements/memory/arm/68.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 57.37 per second) + +11264.00 MiB transferred (58744.38 MiB/sec) + + +General statistics: + total time: 0.1915s + total number of events: 11 + +Latency (ms): + min: 173.20 + avg: 187.52 + max: 190.94 + 95th percentile: 0.00 + sum: 2062.71 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1875/0.01 + diff --git a/measurements/memory/arm/69.log b/measurements/memory/arm/69.log new file mode 100644 index 0000000..b447b86 --- /dev/null +++ b/measurements/memory/arm/69.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 83.55 per second) + +11264.00 MiB transferred (85555.65 MiB/sec) + + +General statistics: + total time: 0.1314s + total number of events: 11 + +Latency (ms): + min: 128.61 + avg: 130.16 + max: 131.31 + 95th percentile: 0.00 + sum: 1431.70 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1302/0.00 + diff --git a/measurements/memory/arm/7.log b/measurements/memory/arm/7.log new file mode 100644 index 0000000..c5f8112 --- /dev/null +++ b/measurements/memory/arm/7.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 70.18 per second) + +11264.00 MiB transferred (71861.48 MiB/sec) + + +General statistics: + total time: 0.1564s + total number of events: 11 + +Latency (ms): + min: 148.34 + avg: 151.85 + max: 156.39 + 95th percentile: 0.00 + sum: 1670.31 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1518/0.00 + diff --git a/measurements/memory/arm/70.log b/measurements/memory/arm/70.log new file mode 100644 index 0000000..dcdab2c --- /dev/null +++ b/measurements/memory/arm/70.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 54.73 per second) + +11264.00 MiB transferred (56043.43 MiB/sec) + + +General statistics: + total time: 0.2006s + total number of events: 11 + +Latency (ms): + min: 196.41 + avg: 198.56 + max: 200.48 + 95th percentile: 0.00 + sum: 2184.15 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1986/0.00 + diff --git a/measurements/memory/arm/71.log b/measurements/memory/arm/71.log new file mode 100644 index 0000000..48628c3 --- /dev/null +++ b/measurements/memory/arm/71.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 46.94 per second) + +11264.00 MiB transferred (48064.04 MiB/sec) + + +General statistics: + total time: 0.2340s + total number of events: 11 + +Latency (ms): + min: 217.66 + avg: 225.70 + max: 233.82 + 95th percentile: 0.00 + sum: 2482.68 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.2257/0.01 + diff --git a/measurements/memory/arm/72.log b/measurements/memory/arm/72.log new file mode 100644 index 0000000..b0d2c4b --- /dev/null +++ b/measurements/memory/arm/72.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 32.11 per second) + +11264.00 MiB transferred (32879.24 MiB/sec) + + +General statistics: + total time: 0.3423s + total number of events: 11 + +Latency (ms): + min: 341.75 + avg: 341.82 + max: 342.05 + 95th percentile: 0.00 + sum: 3760.05 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.3418/0.00 + diff --git a/measurements/memory/arm/73.log b/measurements/memory/arm/73.log new file mode 100644 index 0000000..875c76e --- /dev/null +++ b/measurements/memory/arm/73.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 95.19 per second) + +11264.00 MiB transferred (97477.39 MiB/sec) + + +General statistics: + total time: 0.1153s + total number of events: 11 + +Latency (ms): + min: 101.47 + avg: 106.85 + max: 111.15 + 95th percentile: 0.00 + sum: 1175.34 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1068/0.00 + diff --git a/measurements/memory/arm/74.log b/measurements/memory/arm/74.log new file mode 100644 index 0000000..77e0da0 --- /dev/null +++ b/measurements/memory/arm/74.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 79.41 per second) + +11264.00 MiB transferred (81319.13 MiB/sec) + + +General statistics: + total time: 0.1382s + total number of events: 11 + +Latency (ms): + min: 110.77 + avg: 124.60 + max: 130.74 + 95th percentile: 0.00 + sum: 1370.65 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1246/0.01 + diff --git a/measurements/memory/arm/75.log b/measurements/memory/arm/75.log new file mode 100644 index 0000000..1f0f34c --- /dev/null +++ b/measurements/memory/arm/75.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 78.73 per second) + +11264.00 MiB transferred (80617.80 MiB/sec) + + +General statistics: + total time: 0.1394s + total number of events: 11 + +Latency (ms): + min: 128.04 + avg: 137.86 + max: 139.37 + 95th percentile: 0.00 + sum: 1516.44 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1379/0.00 + diff --git a/measurements/memory/arm/76.log b/measurements/memory/arm/76.log new file mode 100644 index 0000000..287bbf3 --- /dev/null +++ b/measurements/memory/arm/76.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 95.71 per second) + +11264.00 MiB transferred (98002.37 MiB/sec) + + +General statistics: + total time: 0.1146s + total number of events: 11 + +Latency (ms): + min: 99.41 + avg: 102.56 + max: 107.63 + 95th percentile: 0.00 + sum: 1128.11 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1026/0.00 + diff --git a/measurements/memory/arm/77.log b/measurements/memory/arm/77.log new file mode 100644 index 0000000..f7cbd4d --- /dev/null +++ b/measurements/memory/arm/77.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 66.60 per second) + +11264.00 MiB transferred (68196.40 MiB/sec) + + +General statistics: + total time: 0.1648s + total number of events: 11 + +Latency (ms): + min: 146.58 + avg: 160.22 + max: 164.09 + 95th percentile: 0.00 + sum: 1762.37 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1602/0.01 + diff --git a/measurements/memory/arm/78.log b/measurements/memory/arm/78.log new file mode 100644 index 0000000..21df507 --- /dev/null +++ b/measurements/memory/arm/78.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 61.64 per second) + +11264.00 MiB transferred (63117.08 MiB/sec) + + +General statistics: + total time: 0.1782s + total number of events: 11 + +Latency (ms): + min: 176.50 + avg: 177.50 + max: 178.10 + 95th percentile: 0.00 + sum: 1952.50 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1775/0.00 + diff --git a/measurements/memory/arm/79.log b/measurements/memory/arm/79.log new file mode 100644 index 0000000..74cf61b --- /dev/null +++ b/measurements/memory/arm/79.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 87.76 per second) + +11264.00 MiB transferred (89871.15 MiB/sec) + + +General statistics: + total time: 0.1250s + total number of events: 11 + +Latency (ms): + min: 116.75 + avg: 119.35 + max: 124.91 + 95th percentile: 0.00 + sum: 1312.85 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1194/0.00 + diff --git a/measurements/memory/arm/8.log b/measurements/memory/arm/8.log new file mode 100644 index 0000000..20a3182 --- /dev/null +++ b/measurements/memory/arm/8.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 64.63 per second) + +11264.00 MiB transferred (66182.91 MiB/sec) + + +General statistics: + total time: 0.1699s + total number of events: 11 + +Latency (ms): + min: 169.37 + avg: 169.59 + max: 169.85 + 95th percentile: 0.00 + sum: 1865.51 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1696/0.00 + diff --git a/measurements/memory/arm/80.log b/measurements/memory/arm/80.log new file mode 100644 index 0000000..41c3fb0 --- /dev/null +++ b/measurements/memory/arm/80.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 57.68 per second) + +11264.00 MiB transferred (59065.98 MiB/sec) + + +General statistics: + total time: 0.1904s + total number of events: 11 + +Latency (ms): + min: 176.61 + avg: 182.49 + max: 189.33 + 95th percentile: 0.00 + sum: 2007.39 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1825/0.00 + diff --git a/measurements/memory/arm/81.log b/measurements/memory/arm/81.log new file mode 100644 index 0000000..c28366c --- /dev/null +++ b/measurements/memory/arm/81.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 71.57 per second) + +11264.00 MiB transferred (73290.39 MiB/sec) + + +General statistics: + total time: 0.1534s + total number of events: 11 + +Latency (ms): + min: 148.55 + avg: 151.39 + max: 153.34 + 95th percentile: 0.00 + sum: 1665.33 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1514/0.00 + diff --git a/measurements/memory/arm/82.log b/measurements/memory/arm/82.log new file mode 100644 index 0000000..cd8f58e --- /dev/null +++ b/measurements/memory/arm/82.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 66.57 per second) + +11264.00 MiB transferred (68164.63 MiB/sec) + + +General statistics: + total time: 0.1650s + total number of events: 11 + +Latency (ms): + min: 164.70 + avg: 164.76 + max: 164.80 + 95th percentile: 0.00 + sum: 1812.33 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1648/0.00 + diff --git a/measurements/memory/arm/83.log b/measurements/memory/arm/83.log new file mode 100644 index 0000000..4acba68 --- /dev/null +++ b/measurements/memory/arm/83.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 62.44 per second) + +11264.00 MiB transferred (63940.42 MiB/sec) + + +General statistics: + total time: 0.1759s + total number of events: 11 + +Latency (ms): + min: 172.04 + avg: 175.14 + max: 175.81 + 95th percentile: 0.00 + sum: 1926.60 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1751/0.00 + diff --git a/measurements/memory/arm/84.log b/measurements/memory/arm/84.log new file mode 100644 index 0000000..f6b87e9 --- /dev/null +++ b/measurements/memory/arm/84.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 55.99 per second) + +11264.00 MiB transferred (57328.70 MiB/sec) + + +General statistics: + total time: 0.1962s + total number of events: 11 + +Latency (ms): + min: 194.24 + avg: 195.00 + max: 196.16 + 95th percentile: 0.00 + sum: 2144.97 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1950/0.00 + diff --git a/measurements/memory/arm/85.log b/measurements/memory/arm/85.log new file mode 100644 index 0000000..ef1f945 --- /dev/null +++ b/measurements/memory/arm/85.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 55.31 per second) + +11264.00 MiB transferred (56635.46 MiB/sec) + + +General statistics: + total time: 0.1985s + total number of events: 11 + +Latency (ms): + min: 195.61 + avg: 197.50 + max: 198.29 + 95th percentile: 0.00 + sum: 2172.50 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1975/0.00 + diff --git a/measurements/memory/arm/86.log b/measurements/memory/arm/86.log new file mode 100644 index 0000000..b5afd86 --- /dev/null +++ b/measurements/memory/arm/86.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 63.29 per second) + +11264.00 MiB transferred (64804.91 MiB/sec) + + +General statistics: + total time: 0.1735s + total number of events: 11 + +Latency (ms): + min: 168.29 + avg: 171.71 + max: 173.45 + 95th percentile: 0.00 + sum: 1888.76 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1717/0.00 + diff --git a/measurements/memory/arm/87.log b/measurements/memory/arm/87.log new file mode 100644 index 0000000..0e810be --- /dev/null +++ b/measurements/memory/arm/87.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 64.20 per second) + +11264.00 MiB transferred (65738.33 MiB/sec) + + +General statistics: + total time: 0.1711s + total number of events: 11 + +Latency (ms): + min: 169.39 + avg: 170.63 + max: 170.99 + 95th percentile: 0.00 + sum: 1876.90 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1706/0.00 + diff --git a/measurements/memory/arm/88.log b/measurements/memory/arm/88.log new file mode 100644 index 0000000..c0ba653 --- /dev/null +++ b/measurements/memory/arm/88.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 86.08 per second) + +11264.00 MiB transferred (88142.54 MiB/sec) + + +General statistics: + total time: 0.1275s + total number of events: 11 + +Latency (ms): + min: 123.11 + avg: 124.11 + max: 127.41 + 95th percentile: 0.00 + sum: 1365.21 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1241/0.00 + diff --git a/measurements/memory/arm/89.log b/measurements/memory/arm/89.log new file mode 100644 index 0000000..06b75af --- /dev/null +++ b/measurements/memory/arm/89.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 63.81 per second) + +11264.00 MiB transferred (65339.08 MiB/sec) + + +General statistics: + total time: 0.1721s + total number of events: 11 + +Latency (ms): + min: 158.27 + avg: 166.13 + max: 172.05 + 95th percentile: 0.00 + sum: 1827.47 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1661/0.00 + diff --git a/measurements/memory/arm/9.log b/measurements/memory/arm/9.log new file mode 100644 index 0000000..cf7a226 --- /dev/null +++ b/measurements/memory/arm/9.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 79.14 per second) + +11264.00 MiB transferred (81037.72 MiB/sec) + + +General statistics: + total time: 0.1387s + total number of events: 11 + +Latency (ms): + min: 129.46 + avg: 131.05 + max: 138.56 + 95th percentile: 0.00 + sum: 1441.56 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1311/0.00 + diff --git a/measurements/memory/arm/90.log b/measurements/memory/arm/90.log new file mode 100644 index 0000000..3cf9cef --- /dev/null +++ b/measurements/memory/arm/90.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 70.34 per second) + +11264.00 MiB transferred (72027.37 MiB/sec) + + +General statistics: + total time: 0.1560s + total number of events: 11 + +Latency (ms): + min: 155.73 + avg: 155.80 + max: 155.85 + 95th percentile: 0.00 + sum: 1713.76 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1558/0.00 + diff --git a/measurements/memory/arm/91.log b/measurements/memory/arm/91.log new file mode 100644 index 0000000..a885635 --- /dev/null +++ b/measurements/memory/arm/91.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 65.61 per second) + +11264.00 MiB transferred (67180.78 MiB/sec) + + +General statistics: + total time: 0.1674s + total number of events: 11 + +Latency (ms): + min: 155.33 + avg: 160.09 + max: 167.25 + 95th percentile: 0.00 + sum: 1761.03 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1601/0.00 + diff --git a/measurements/memory/arm/92.log b/measurements/memory/arm/92.log new file mode 100644 index 0000000..20f10a1 --- /dev/null +++ b/measurements/memory/arm/92.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 61.39 per second) + +11264.00 MiB transferred (62858.55 MiB/sec) + + +General statistics: + total time: 0.1789s + total number of events: 11 + +Latency (ms): + min: 168.40 + avg: 173.31 + max: 175.34 + 95th percentile: 0.00 + sum: 1906.39 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1733/0.00 + diff --git a/measurements/memory/arm/93.log b/measurements/memory/arm/93.log new file mode 100644 index 0000000..7389678 --- /dev/null +++ b/measurements/memory/arm/93.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 59.12 per second) + +11264.00 MiB transferred (60536.03 MiB/sec) + + +General statistics: + total time: 0.1858s + total number of events: 11 + +Latency (ms): + min: 179.16 + avg: 183.09 + max: 185.74 + 95th percentile: 0.00 + sum: 2014.02 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1831/0.00 + diff --git a/measurements/memory/arm/94.log b/measurements/memory/arm/94.log new file mode 100644 index 0000000..e47c3a9 --- /dev/null +++ b/measurements/memory/arm/94.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 64.92 per second) + +11264.00 MiB transferred (66480.95 MiB/sec) + + +General statistics: + total time: 0.1691s + total number of events: 11 + +Latency (ms): + min: 159.79 + avg: 162.79 + max: 167.33 + 95th percentile: 0.00 + sum: 1790.66 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1628/0.00 + diff --git a/measurements/memory/arm/95.log b/measurements/memory/arm/95.log new file mode 100644 index 0000000..0b7b72a --- /dev/null +++ b/measurements/memory/arm/95.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 68.98 per second) + +11264.00 MiB transferred (70637.08 MiB/sec) + + +General statistics: + total time: 0.1592s + total number of events: 11 + +Latency (ms): + min: 154.33 + avg: 154.95 + max: 159.09 + 95th percentile: 0.00 + sum: 1704.41 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1549/0.00 + diff --git a/measurements/memory/arm/96.log b/measurements/memory/arm/96.log new file mode 100644 index 0000000..0f99042 --- /dev/null +++ b/measurements/memory/arm/96.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 76.89 per second) + +11264.00 MiB transferred (78735.64 MiB/sec) + + +General statistics: + total time: 0.1428s + total number of events: 11 + +Latency (ms): + min: 135.73 + avg: 138.71 + max: 142.61 + 95th percentile: 0.00 + sum: 1525.78 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1387/0.00 + diff --git a/measurements/memory/arm/97.log b/measurements/memory/arm/97.log new file mode 100644 index 0000000..2870d6c --- /dev/null +++ b/measurements/memory/arm/97.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 67.05 per second) + +11264.00 MiB transferred (68658.23 MiB/sec) + + +General statistics: + total time: 0.1638s + total number of events: 11 + +Latency (ms): + min: 151.63 + avg: 155.11 + max: 160.70 + 95th percentile: 0.00 + sum: 1706.19 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1551/0.00 + diff --git a/measurements/memory/arm/98.log b/measurements/memory/arm/98.log new file mode 100644 index 0000000..01bef8c --- /dev/null +++ b/measurements/memory/arm/98.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 75.82 per second) + +11264.00 MiB transferred (77643.13 MiB/sec) + + +General statistics: + total time: 0.1448s + total number of events: 11 + +Latency (ms): + min: 144.20 + avg: 144.37 + max: 144.68 + 95th percentile: 0.00 + sum: 1588.03 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1444/0.00 + diff --git a/measurements/memory/arm/99.log b/measurements/memory/arm/99.log new file mode 100644 index 0000000..ad22ac8 --- /dev/null +++ b/measurements/memory/arm/99.log @@ -0,0 +1,37 @@ +sysbench 1.0.20 (using system LuaJIT 2.1.1753364724) + +Running the test with following options: +Number of threads: 11 +Initializing random number generator from current time + + +Running memory speed test with the following options: + block size: 1048576KiB + total size: 20480MiB + operation: read + scope: global + +Initializing worker threads... + +Threads started! + +Total operations: 11 ( 84.27 per second) + +11264.00 MiB transferred (86292.36 MiB/sec) + + +General statistics: + total time: 0.1303s + total number of events: 11 + +Latency (ms): + min: 126.00 + avg: 127.01 + max: 130.19 + 95th percentile: 0.00 + sum: 1397.06 + +Threads fairness: + events (avg/stddev): 1.0000/0.00 + execution time (avg/stddev): 0.1270/0.00 + diff --git a/measurements/variant_baseline/arm/1.log b/measurements/variant_baseline/arm/1.log index 9f2eb9c..3f4b18b 100644 --- a/measurements/variant_baseline/arm/1.log +++ b/measurements/variant_baseline/arm/1.log @@ -5,48 +5,54 @@ W(n) per iteration: 183700.00 flops Q(n) per iteration: 6512 bytes I(n): 28.21 flops/byte W(n): 36740000.00 flops -T(n): 18154426.88 cycles -P(n): 2.02 flops/cycle +T(n): 18519913.62 cycles +P(n): 1.98 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 727400.00 flops Q(n) per iteration: 12912 bytes I(n): 56.34 flops/byte W(n): 145480000.00 flops -T(n): 74812980.50 cycles -P(n): 1.94 flops/cycle +T(n): 76255697.50 cycles +P(n): 1.91 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 2894800.00 flops Q(n) per iteration: 25712 bytes I(n): 112.59 flops/byte W(n): 578960000.00 flops -T(n): 322910201.00 cycles -P(n): 1.79 flops/cycle +T(n): 318362714.00 cycles +P(n): 1.82 flops/cycle === N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 2235914629.00 cycles +P(n): 1.61 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 72074000.00 flops Q(n) per iteration: 128112 bytes I(n): 562.59 flops/byte W(n): 14414800000.00 flops -T(n): 9012317356.00 cycles -P(n): 1.60 flops/cycle +T(n): 9387881894.00 cycles +P(n): 1.54 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 288148000.00 flops Q(n) per iteration: 256112 bytes I(n): 1125.09 flops/byte W(n): 57629600000.00 flops -T(n): 37090810409.00 cycles -P(n): 1.55 flops/cycle +T(n): 37672272939.00 cycles +P(n): 1.53 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 1800370000.00 flops Q(n) per iteration: 640112 bytes I(n): 2812.59 flops/byte W(n): 360074000000.00 flops -T(n): 234807270918.00 cycles -P(n): 1.53 flops/cycle +T(n): 234323381780.00 cycles +P(n): 1.54 flops/cycle diff --git a/measurements/variant_baseline/arm/2.log b/measurements/variant_baseline/arm/2.log new file mode 100644 index 0000000..85686eb --- /dev/null +++ b/measurements/variant_baseline/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_baseline.cpp -o main_baseline -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 18091961.62 cycles +P(n): 2.03 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 74904528.00 cycles +P(n): 1.94 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 323946225.00 cycles +P(n): 1.79 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 2220373486.00 cycles +P(n): 1.62 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 9097774774.00 cycles +P(n): 1.58 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 37320008701.00 cycles +P(n): 1.54 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 236995763096.00 cycles +P(n): 1.52 flops/cycle + diff --git a/measurements/variant_baseline/arm/3.log b/measurements/variant_baseline/arm/3.log new file mode 100644 index 0000000..c44eedd --- /dev/null +++ b/measurements/variant_baseline/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_baseline.cpp -o main_baseline -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 18269570.62 cycles +P(n): 2.01 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 75215747.00 cycles +P(n): 1.93 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 319362431.00 cycles +P(n): 1.81 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 2198822242.00 cycles +P(n): 1.64 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 8980802154.00 cycles +P(n): 1.61 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 36998543564.00 cycles +P(n): 1.56 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 235915777026.00 cycles +P(n): 1.53 flops/cycle + diff --git a/measurements/variant_branchless/arm/1.log b/measurements/variant_branchless/arm/1.log index c6ae4b2..be32c6f 100644 --- a/measurements/variant_branchless/arm/1.log +++ b/measurements/variant_branchless/arm/1.log @@ -5,54 +5,54 @@ W(n) per iteration: 183700.00 flops Q(n) per iteration: 6512 bytes I(n): 28.21 flops/byte W(n): 36740000.00 flops -T(n): 16997520.38 cycles -P(n): 2.16 flops/cycle +T(n): 17074790.38 cycles +P(n): 2.15 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 727400.00 flops Q(n) per iteration: 12912 bytes I(n): 56.34 flops/byte W(n): 145480000.00 flops -T(n): 67764539.00 cycles -P(n): 2.15 flops/cycle +T(n): 67448874.50 cycles +P(n): 2.16 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 2894800.00 flops Q(n) per iteration: 25712 bytes I(n): 112.59 flops/byte W(n): 578960000.00 flops -T(n): 269251877.00 cycles -P(n): 2.15 flops/cycle +T(n): 266691861.00 cycles +P(n): 2.17 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 18037000.00 flops Q(n) per iteration: 64112 bytes I(n): 281.34 flops/byte W(n): 3607400000.00 flops -T(n): 1667878996.00 cycles -P(n): 2.16 flops/cycle +T(n): 1661210085.00 cycles +P(n): 2.17 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 72074000.00 flops Q(n) per iteration: 128112 bytes I(n): 562.59 flops/byte W(n): 14414800000.00 flops -T(n): 6785785070.00 cycles -P(n): 2.12 flops/cycle +T(n): 6756280473.00 cycles +P(n): 2.13 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 288148000.00 flops Q(n) per iteration: 256112 bytes I(n): 1125.09 flops/byte W(n): 57629600000.00 flops -T(n): 26925708355.00 cycles -P(n): 2.14 flops/cycle +T(n): 27171263293.00 cycles +P(n): 2.12 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 1800370000.00 flops Q(n) per iteration: 640112 bytes I(n): 2812.59 flops/byte W(n): 360074000000.00 flops -T(n): 167430181147.00 cycles -P(n): 2.15 flops/cycle +T(n): 164723372233.00 cycles +P(n): 2.19 flops/cycle diff --git a/measurements/variant_branchless/arm/2.log b/measurements/variant_branchless/arm/2.log new file mode 100644 index 0000000..e420944 --- /dev/null +++ b/measurements/variant_branchless/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_branchless.cpp -o main_branchless -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 17032592.62 cycles +P(n): 2.16 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 66853787.50 cycles +P(n): 2.18 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 264804936.00 cycles +P(n): 2.19 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 1655724217.00 cycles +P(n): 2.18 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 6695963590.00 cycles +P(n): 2.15 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 26555878223.00 cycles +P(n): 2.17 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 165943163628.00 cycles +P(n): 2.17 flops/cycle + diff --git a/measurements/variant_branchless/arm/3.log b/measurements/variant_branchless/arm/3.log new file mode 100644 index 0000000..0f10083 --- /dev/null +++ b/measurements/variant_branchless/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_branchless.cpp -o main_branchless -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 17006998.38 cycles +P(n): 2.16 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 67065002.50 cycles +P(n): 2.17 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 265223905.00 cycles +P(n): 2.18 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 1653007159.00 cycles +P(n): 2.18 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 6740132027.00 cycles +P(n): 2.14 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 26592000772.00 cycles +P(n): 2.17 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 166171547345.00 cycles +P(n): 2.17 flops/cycle + diff --git a/measurements/variant_branchless_sqrtfree/arm/1.log b/measurements/variant_branchless_sqrtfree/arm/1.log index c0a17c2..5236dfc 100644 --- a/measurements/variant_branchless_sqrtfree/arm/1.log +++ b/measurements/variant_branchless_sqrtfree/arm/1.log @@ -5,54 +5,54 @@ W(n) per iteration: 183700.00 flops Q(n) per iteration: 6512 bytes I(n): 28.21 flops/byte W(n): 36740000.00 flops -T(n): 17385705.00 cycles -P(n): 2.11 flops/cycle +T(n): 16994847.88 cycles +P(n): 2.16 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 727400.00 flops Q(n) per iteration: 12912 bytes I(n): 56.34 flops/byte W(n): 145480000.00 flops -T(n): 66976913.50 cycles -P(n): 2.17 flops/cycle +T(n): 66600443.50 cycles +P(n): 2.18 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 2894800.00 flops Q(n) per iteration: 25712 bytes I(n): 112.59 flops/byte W(n): 578960000.00 flops -T(n): 266594191.00 cycles -P(n): 2.17 flops/cycle +T(n): 263642655.00 cycles +P(n): 2.20 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 18037000.00 flops Q(n) per iteration: 64112 bytes I(n): 281.34 flops/byte W(n): 3607400000.00 flops -T(n): 1659465829.00 cycles -P(n): 2.17 flops/cycle +T(n): 1654799864.00 cycles +P(n): 2.18 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 72074000.00 flops Q(n) per iteration: 128112 bytes I(n): 562.59 flops/byte W(n): 14414800000.00 flops -T(n): 6723368454.00 cycles -P(n): 2.14 flops/cycle +T(n): 6719174873.00 cycles +P(n): 2.15 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 288148000.00 flops Q(n) per iteration: 256112 bytes I(n): 1125.09 flops/byte W(n): 57629600000.00 flops -T(n): 26939978394.00 cycles -P(n): 2.14 flops/cycle +T(n): 26585178135.00 cycles +P(n): 2.17 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 1800370000.00 flops Q(n) per iteration: 640112 bytes I(n): 2812.59 flops/byte W(n): 360074000000.00 flops -T(n): 166305320990.00 cycles -P(n): 2.17 flops/cycle +T(n): 166588550172.00 cycles +P(n): 2.16 flops/cycle diff --git a/measurements/variant_branchless_sqrtfree/arm/2.log b/measurements/variant_branchless_sqrtfree/arm/2.log new file mode 100644 index 0000000..6cd51bd --- /dev/null +++ b/measurements/variant_branchless_sqrtfree/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_branchless_sqrtfree.cpp -o main_branchless_sqrtfree -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 16939009.25 cycles +P(n): 2.17 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 66297975.50 cycles +P(n): 2.19 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 263376267.00 cycles +P(n): 2.20 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 1637432815.00 cycles +P(n): 2.20 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 6657953874.00 cycles +P(n): 2.17 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 26475624015.00 cycles +P(n): 2.18 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 165290358959.00 cycles +P(n): 2.18 flops/cycle + diff --git a/measurements/variant_branchless_sqrtfree/arm/3.log b/measurements/variant_branchless_sqrtfree/arm/3.log new file mode 100644 index 0000000..07d103b --- /dev/null +++ b/measurements/variant_branchless_sqrtfree/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_branchless_sqrtfree.cpp -o main_branchless_sqrtfree -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 16925943.12 cycles +P(n): 2.17 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 66346832.50 cycles +P(n): 2.19 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 264215591.00 cycles +P(n): 2.19 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 1636514007.00 cycles +P(n): 2.20 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 6678250320.00 cycles +P(n): 2.16 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 26550461709.00 cycles +P(n): 2.17 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 164730393682.00 cycles +P(n): 2.19 flops/cycle + diff --git a/measurements/variant_cutoff/arm/1.log b/measurements/variant_cutoff/arm/1.log index 4f44c0d..8537605 100644 --- a/measurements/variant_cutoff/arm/1.log +++ b/measurements/variant_cutoff/arm/1.log @@ -5,54 +5,54 @@ W(n) per iteration: 183700.00 flops Q(n) per iteration: 6512 bytes I(n): 28.21 flops/byte W(n): 36740000.00 flops -T(n): 8713477.81 cycles -P(n): 4.22 flops/cycle +T(n): 8791913.88 cycles +P(n): 4.18 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 727400.00 flops Q(n) per iteration: 12912 bytes I(n): 56.34 flops/byte W(n): 145480000.00 flops -T(n): 40792521.50 cycles -P(n): 3.57 flops/cycle +T(n): 41158496.75 cycles +P(n): 3.53 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 2894800.00 flops Q(n) per iteration: 25712 bytes I(n): 112.59 flops/byte W(n): 578960000.00 flops -T(n): 132995997.00 cycles -P(n): 4.35 flops/cycle +T(n): 133925307.00 cycles +P(n): 4.32 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 18037000.00 flops Q(n) per iteration: 64112 bytes I(n): 281.34 flops/byte W(n): 3607400000.00 flops -T(n): 813430901.00 cycles -P(n): 4.43 flops/cycle +T(n): 823137794.00 cycles +P(n): 4.38 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 72074000.00 flops Q(n) per iteration: 128112 bytes I(n): 562.59 flops/byte W(n): 14414800000.00 flops -T(n): 3155176718.00 cycles -P(n): 4.57 flops/cycle +T(n): 3167884952.00 cycles +P(n): 4.55 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 288148000.00 flops Q(n) per iteration: 256112 bytes I(n): 1125.09 flops/byte W(n): 57629600000.00 flops -T(n): 12638893219.00 cycles -P(n): 4.56 flops/cycle +T(n): 12582829866.00 cycles +P(n): 4.58 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 1800370000.00 flops Q(n) per iteration: 640112 bytes I(n): 2812.59 flops/byte W(n): 360074000000.00 flops -T(n): 76387953469.00 cycles -P(n): 4.71 flops/cycle +T(n): 75970386305.00 cycles +P(n): 4.74 flops/cycle diff --git a/measurements/variant_cutoff/arm/2.log b/measurements/variant_cutoff/arm/2.log new file mode 100644 index 0000000..58d5af4 --- /dev/null +++ b/measurements/variant_cutoff/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_cutoff.cpp -o main_cutoff -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 8789882.25 cycles +P(n): 4.18 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 41157310.00 cycles +P(n): 3.53 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 134011989.00 cycles +P(n): 4.32 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 821974719.00 cycles +P(n): 4.39 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 3180457398.00 cycles +P(n): 4.53 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 12572786176.00 cycles +P(n): 4.58 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 76601440906.00 cycles +P(n): 4.70 flops/cycle + diff --git a/measurements/variant_cutoff/arm/3.log b/measurements/variant_cutoff/arm/3.log new file mode 100644 index 0000000..3702999 --- /dev/null +++ b/measurements/variant_cutoff/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_cutoff.cpp -o main_cutoff -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 8791731.94 cycles +P(n): 4.18 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 41240033.75 cycles +P(n): 3.53 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 134229626.00 cycles +P(n): 4.31 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 819718803.00 cycles +P(n): 4.40 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 3157597976.00 cycles +P(n): 4.57 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 12581653566.00 cycles +P(n): 4.58 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 76309138275.00 cycles +P(n): 4.72 flops/cycle + diff --git a/measurements/variant_cutoff_callsite/arm/1.log b/measurements/variant_cutoff_callsite/arm/1.log index 0fa0e11..2341faf 100644 --- a/measurements/variant_cutoff_callsite/arm/1.log +++ b/measurements/variant_cutoff_callsite/arm/1.log @@ -5,7 +5,7 @@ W(n) per iteration: 183700.00 flops Q(n) per iteration: 6512 bytes I(n): 28.21 flops/byte W(n): 36740000.00 flops -T(n): 9210738.56 cycles +T(n): 9210543.88 cycles P(n): 3.99 flops/cycle === N_P=100 N_ITER=200 === @@ -13,7 +13,7 @@ W(n) per iteration: 727400.00 flops Q(n) per iteration: 12912 bytes I(n): 56.34 flops/byte W(n): 145480000.00 flops -T(n): 43525263.50 cycles +T(n): 43547204.50 cycles P(n): 3.34 flops/cycle === N_P=200 N_ITER=200 === @@ -21,38 +21,38 @@ W(n) per iteration: 2894800.00 flops Q(n) per iteration: 25712 bytes I(n): 112.59 flops/byte W(n): 578960000.00 flops -T(n): 141467337.00 cycles -P(n): 4.09 flops/cycle +T(n): 141046620.00 cycles +P(n): 4.10 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 18037000.00 flops Q(n) per iteration: 64112 bytes I(n): 281.34 flops/byte W(n): 3607400000.00 flops -T(n): 870479259.00 cycles -P(n): 4.14 flops/cycle +T(n): 864905902.00 cycles +P(n): 4.17 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 72074000.00 flops Q(n) per iteration: 128112 bytes I(n): 562.59 flops/byte W(n): 14414800000.00 flops -T(n): 3340600189.00 cycles -P(n): 4.32 flops/cycle +T(n): 3329136423.00 cycles +P(n): 4.33 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 288148000.00 flops Q(n) per iteration: 256112 bytes I(n): 1125.09 flops/byte W(n): 57629600000.00 flops -T(n): 13325615778.00 cycles -P(n): 4.32 flops/cycle +T(n): 13269871529.00 cycles +P(n): 4.34 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 1800370000.00 flops Q(n) per iteration: 640112 bytes I(n): 2812.59 flops/byte W(n): 360074000000.00 flops -T(n): 79881699571.00 cycles -P(n): 4.51 flops/cycle +T(n): 79388198061.00 cycles +P(n): 4.54 flops/cycle diff --git a/measurements/variant_cutoff_callsite/arm/2.log b/measurements/variant_cutoff_callsite/arm/2.log new file mode 100644 index 0000000..76a2cf1 --- /dev/null +++ b/measurements/variant_cutoff_callsite/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_cutoff_callsite.cpp -o main_cutoff_callsite -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 9210916.88 cycles +P(n): 3.99 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 43577670.25 cycles +P(n): 3.34 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 141469488.00 cycles +P(n): 4.09 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 866035209.00 cycles +P(n): 4.17 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 3352259719.00 cycles +P(n): 4.30 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 13395601028.00 cycles +P(n): 4.30 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 79459764831.00 cycles +P(n): 4.53 flops/cycle + diff --git a/measurements/variant_cutoff_callsite/arm/3.log b/measurements/variant_cutoff_callsite/arm/3.log new file mode 100644 index 0000000..0e2a5eb --- /dev/null +++ b/measurements/variant_cutoff_callsite/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_cutoff_callsite.cpp -o main_cutoff_callsite -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 9221152.25 cycles +P(n): 3.98 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 43576456.75 cycles +P(n): 3.34 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 141908818.00 cycles +P(n): 4.08 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 867445885.00 cycles +P(n): 4.16 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 3340595294.00 cycles +P(n): 4.32 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 13419490551.00 cycles +P(n): 4.29 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 79764962547.00 cycles +P(n): 4.51 flops/cycle + diff --git a/measurements/variant_float_baseline/arm/1.log b/measurements/variant_float_baseline/arm/1.log index f6bda9e..e762031 100644 --- a/measurements/variant_float_baseline/arm/1.log +++ b/measurements/variant_float_baseline/arm/1.log @@ -5,54 +5,54 @@ W(n) per iteration: 183700.00 flops Q(n) per iteration: 3256 bytes I(n): 56.42 flops/byte W(n): 36740000.00 flops -T(n): 12278939.00 cycles -P(n): 2.99 flops/cycle +T(n): 12372168.00 cycles +P(n): 2.97 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 727400.00 flops Q(n) per iteration: 6456 bytes I(n): 112.67 flops/byte W(n): 145480000.00 flops -T(n): 49596532.00 cycles -P(n): 2.93 flops/cycle +T(n): 51675360.00 cycles +P(n): 2.82 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 2894800.00 flops Q(n) per iteration: 12856 bytes I(n): 225.17 flops/byte W(n): 578960000.00 flops -T(n): 203275376.00 cycles -P(n): 2.85 flops/cycle +T(n): 202560832.00 cycles +P(n): 2.86 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 18037000.00 flops Q(n) per iteration: 32056 bytes I(n): 562.67 flops/byte W(n): 3607399936.00 flops -T(n): 1414936832.00 cycles -P(n): 2.55 flops/cycle +T(n): 1397319552.00 cycles +P(n): 2.58 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 72074000.00 flops Q(n) per iteration: 64056 bytes I(n): 1125.17 flops/byte W(n): 14414799872.00 flops -T(n): 6075253248.00 cycles -P(n): 2.37 flops/cycle +T(n): 6051383808.00 cycles +P(n): 2.38 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 288148000.00 flops Q(n) per iteration: 128056 bytes I(n): 2250.17 flops/byte W(n): 57629601792.00 flops -T(n): 26570563584.00 cycles -P(n): 2.17 flops/cycle +T(n): 26467629056.00 cycles +P(n): 2.18 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 1800369920.00 flops Q(n) per iteration: 320056 bytes I(n): 5625.17 flops/byte W(n): 360073986048.00 flops -T(n): 194256240640.00 cycles -P(n): 1.85 flops/cycle +T(n): 192847380480.00 cycles +P(n): 1.87 flops/cycle diff --git a/measurements/variant_float_baseline/arm/2.log b/measurements/variant_float_baseline/arm/2.log new file mode 100644 index 0000000..61697c0 --- /dev/null +++ b/measurements/variant_float_baseline/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float.cpp ./variants/variant_float_baseline.cpp -o main_float_baseline -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 3256 bytes +I(n): 56.42 flops/byte +W(n): 36740000.00 flops +T(n): 12299523.00 cycles +P(n): 2.99 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 6456 bytes +I(n): 112.67 flops/byte +W(n): 145480000.00 flops +T(n): 51576444.00 cycles +P(n): 2.82 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 12856 bytes +I(n): 225.17 flops/byte +W(n): 578960000.00 flops +T(n): 203345584.00 cycles +P(n): 2.85 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 562.67 flops/byte +W(n): 3607399936.00 flops +T(n): 1412056192.00 cycles +P(n): 2.55 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 1125.17 flops/byte +W(n): 14414799872.00 flops +T(n): 6096742912.00 cycles +P(n): 2.36 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 2250.17 flops/byte +W(n): 57629601792.00 flops +T(n): 26863925248.00 cycles +P(n): 2.15 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800369920.00 flops +Q(n) per iteration: 320056 bytes +I(n): 5625.17 flops/byte +W(n): 360073986048.00 flops +T(n): 194930032640.00 cycles +P(n): 1.85 flops/cycle + diff --git a/measurements/variant_float_baseline/arm/3.log b/measurements/variant_float_baseline/arm/3.log new file mode 100644 index 0000000..109f58e --- /dev/null +++ b/measurements/variant_float_baseline/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float.cpp ./variants/variant_float_baseline.cpp -o main_float_baseline -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 3256 bytes +I(n): 56.42 flops/byte +W(n): 36740000.00 flops +T(n): 12325934.00 cycles +P(n): 2.98 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 6456 bytes +I(n): 112.67 flops/byte +W(n): 145480000.00 flops +T(n): 50132352.00 cycles +P(n): 2.90 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 12856 bytes +I(n): 225.17 flops/byte +W(n): 578960000.00 flops +T(n): 203162832.00 cycles +P(n): 2.85 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 562.67 flops/byte +W(n): 3607399936.00 flops +T(n): 1401285376.00 cycles +P(n): 2.57 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 1125.17 flops/byte +W(n): 14414799872.00 flops +T(n): 6096781312.00 cycles +P(n): 2.36 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 2250.17 flops/byte +W(n): 57629601792.00 flops +T(n): 26538309632.00 cycles +P(n): 2.17 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800369920.00 flops +Q(n) per iteration: 320056 bytes +I(n): 5625.17 flops/byte +W(n): 360073986048.00 flops +T(n): 195303145472.00 cycles +P(n): 1.84 flops/cycle + diff --git a/measurements/variant_float_inlining/arm/1.log b/measurements/variant_float_inlining/arm/1.log index c2c0262..dbdea40 100644 --- a/measurements/variant_float_inlining/arm/1.log +++ b/measurements/variant_float_inlining/arm/1.log @@ -5,15 +5,15 @@ W(n) per iteration: 183700.00 flops Q(n) per iteration: 3256 bytes I(n): 56.42 flops/byte W(n): 36740000.00 flops -T(n): 10775829.00 cycles -P(n): 3.41 flops/cycle +T(n): 10744000.00 cycles +P(n): 3.42 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 727400.00 flops Q(n) per iteration: 6456 bytes I(n): 112.67 flops/byte W(n): 145480000.00 flops -T(n): 47963000.00 cycles +T(n): 47936504.00 cycles P(n): 3.03 flops/cycle === N_P=200 N_ITER=200 === @@ -21,7 +21,7 @@ W(n) per iteration: 2894800.00 flops Q(n) per iteration: 12856 bytes I(n): 225.17 flops/byte W(n): 578960000.00 flops -T(n): 190335120.00 cycles +T(n): 190173216.00 cycles P(n): 3.04 flops/cycle === N_P=500 N_ITER=200 === @@ -29,30 +29,30 @@ W(n) per iteration: 18037000.00 flops Q(n) per iteration: 32056 bytes I(n): 562.67 flops/byte W(n): 3607399936.00 flops -T(n): 1412006400.00 cycles -P(n): 2.55 flops/cycle +T(n): 1404110336.00 cycles +P(n): 2.57 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 72074000.00 flops Q(n) per iteration: 64056 bytes I(n): 1125.17 flops/byte W(n): 14414799872.00 flops -T(n): 6137234944.00 cycles -P(n): 2.35 flops/cycle +T(n): 6120075776.00 cycles +P(n): 2.36 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 288148000.00 flops Q(n) per iteration: 128056 bytes I(n): 2250.17 flops/byte W(n): 57629601792.00 flops -T(n): 27256184832.00 cycles -P(n): 2.11 flops/cycle +T(n): 27074254848.00 cycles +P(n): 2.13 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 1800369920.00 flops Q(n) per iteration: 320056 bytes I(n): 5625.17 flops/byte W(n): 360073986048.00 flops -T(n): 194280062976.00 cycles -P(n): 1.85 flops/cycle +T(n): 192850231296.00 cycles +P(n): 1.87 flops/cycle diff --git a/measurements/variant_float_inlining/arm/2.log b/measurements/variant_float_inlining/arm/2.log new file mode 100644 index 0000000..8f7a2a2 --- /dev/null +++ b/measurements/variant_float_inlining/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float.cpp ./variants/variant_float_inlining.cpp -o main_float_inlining -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 3256 bytes +I(n): 56.42 flops/byte +W(n): 36740000.00 flops +T(n): 10826171.00 cycles +P(n): 3.39 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 6456 bytes +I(n): 112.67 flops/byte +W(n): 145480000.00 flops +T(n): 48929320.00 cycles +P(n): 2.97 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 12856 bytes +I(n): 225.17 flops/byte +W(n): 578960000.00 flops +T(n): 190686384.00 cycles +P(n): 3.04 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 562.67 flops/byte +W(n): 3607399936.00 flops +T(n): 1427274624.00 cycles +P(n): 2.53 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 1125.17 flops/byte +W(n): 14414799872.00 flops +T(n): 6172825600.00 cycles +P(n): 2.34 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 2250.17 flops/byte +W(n): 57629601792.00 flops +T(n): 27468242944.00 cycles +P(n): 2.10 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800369920.00 flops +Q(n) per iteration: 320056 bytes +I(n): 5625.17 flops/byte +W(n): 360073986048.00 flops +T(n): 192917569536.00 cycles +P(n): 1.87 flops/cycle + diff --git a/measurements/variant_float_inlining/arm/3.log b/measurements/variant_float_inlining/arm/3.log new file mode 100644 index 0000000..71ef928 --- /dev/null +++ b/measurements/variant_float_inlining/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float.cpp ./variants/variant_float_inlining.cpp -o main_float_inlining -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 3256 bytes +I(n): 56.42 flops/byte +W(n): 36740000.00 flops +T(n): 10837919.00 cycles +P(n): 3.39 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 6456 bytes +I(n): 112.67 flops/byte +W(n): 145480000.00 flops +T(n): 48233604.00 cycles +P(n): 3.02 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 12856 bytes +I(n): 225.17 flops/byte +W(n): 578960000.00 flops +T(n): 191050976.00 cycles +P(n): 3.03 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 562.67 flops/byte +W(n): 3607399936.00 flops +T(n): 1424769280.00 cycles +P(n): 2.53 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 1125.17 flops/byte +W(n): 14414799872.00 flops +T(n): 6287727104.00 cycles +P(n): 2.29 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 2250.17 flops/byte +W(n): 57629601792.00 flops +T(n): 27565033472.00 cycles +P(n): 2.09 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800369920.00 flops +Q(n) per iteration: 320056 bytes +I(n): 5625.17 flops/byte +W(n): 360073986048.00 flops +T(n): 195009839104.00 cycles +P(n): 1.85 flops/cycle + diff --git a/measurements/variant_float_inlining_SoA/arm/1.log b/measurements/variant_float_inlining_SoA/arm/1.log index d4d7811..149ec4a 100644 --- a/measurements/variant_float_inlining_SoA/arm/1.log +++ b/measurements/variant_float_inlining_SoA/arm/1.log @@ -5,54 +5,54 @@ W(n) per iteration: 183700.00 flops Q(n) per iteration: 3256 bytes I(n): 56.42 flops/byte W(n): 36740000.00 flops -T(n): 11407106.00 cycles -P(n): 3.22 flops/cycle +T(n): 11356540.00 cycles +P(n): 3.24 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 727400.00 flops Q(n) per iteration: 6456 bytes I(n): 112.67 flops/byte W(n): 145480000.00 flops -T(n): 45941640.00 cycles -P(n): 3.17 flops/cycle +T(n): 45488284.00 cycles +P(n): 3.20 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 2894800.00 flops Q(n) per iteration: 12856 bytes I(n): 225.17 flops/byte W(n): 578960000.00 flops -T(n): 175844320.00 cycles -P(n): 3.29 flops/cycle +T(n): 175237216.00 cycles +P(n): 3.30 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 18037000.00 flops Q(n) per iteration: 32056 bytes I(n): 562.67 flops/byte W(n): 3607399936.00 flops -T(n): 1161542784.00 cycles -P(n): 3.11 flops/cycle +T(n): 1146521600.00 cycles +P(n): 3.15 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 72074000.00 flops Q(n) per iteration: 64056 bytes I(n): 1125.17 flops/byte W(n): 14414799872.00 flops -T(n): 4913180160.00 cycles -P(n): 2.93 flops/cycle +T(n): 4827586048.00 cycles +P(n): 2.99 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 288148000.00 flops Q(n) per iteration: 128056 bytes I(n): 2250.17 flops/byte W(n): 57629601792.00 flops -T(n): 21352656896.00 cycles -P(n): 2.70 flops/cycle +T(n): 20706713600.00 cycles +P(n): 2.78 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 1800369920.00 flops Q(n) per iteration: 320056 bytes I(n): 5625.17 flops/byte W(n): 360073986048.00 flops -T(n): 153264947200.00 cycles -P(n): 2.35 flops/cycle +T(n): 150425829376.00 cycles +P(n): 2.39 flops/cycle diff --git a/measurements/variant_float_inlining_SoA/arm/2.log b/measurements/variant_float_inlining_SoA/arm/2.log new file mode 100644 index 0000000..98788d3 --- /dev/null +++ b/measurements/variant_float_inlining_SoA/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float.cpp ./variants/variant_float_inlining_SoA.cpp -o main_float_inlining_SoA -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 3256 bytes +I(n): 56.42 flops/byte +W(n): 36740000.00 flops +T(n): 11386587.00 cycles +P(n): 3.23 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 6456 bytes +I(n): 112.67 flops/byte +W(n): 145480000.00 flops +T(n): 45594080.00 cycles +P(n): 3.19 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 12856 bytes +I(n): 225.17 flops/byte +W(n): 578960000.00 flops +T(n): 175489888.00 cycles +P(n): 3.30 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 562.67 flops/byte +W(n): 3607399936.00 flops +T(n): 1146447744.00 cycles +P(n): 3.15 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 1125.17 flops/byte +W(n): 14414799872.00 flops +T(n): 4937291264.00 cycles +P(n): 2.92 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 2250.17 flops/byte +W(n): 57629601792.00 flops +T(n): 21289326592.00 cycles +P(n): 2.71 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800369920.00 flops +Q(n) per iteration: 320056 bytes +I(n): 5625.17 flops/byte +W(n): 360073986048.00 flops +T(n): 154203766784.00 cycles +P(n): 2.34 flops/cycle + diff --git a/measurements/variant_float_inlining_SoA/arm/3.log b/measurements/variant_float_inlining_SoA/arm/3.log new file mode 100644 index 0000000..22caf95 --- /dev/null +++ b/measurements/variant_float_inlining_SoA/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float.cpp ./variants/variant_float_inlining_SoA.cpp -o main_float_inlining_SoA -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 3256 bytes +I(n): 56.42 flops/byte +W(n): 36740000.00 flops +T(n): 11384260.00 cycles +P(n): 3.23 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 6456 bytes +I(n): 112.67 flops/byte +W(n): 145480000.00 flops +T(n): 45578272.00 cycles +P(n): 3.19 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 12856 bytes +I(n): 225.17 flops/byte +W(n): 578960000.00 flops +T(n): 174984688.00 cycles +P(n): 3.31 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 562.67 flops/byte +W(n): 3607399936.00 flops +T(n): 1151401216.00 cycles +P(n): 3.13 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 1125.17 flops/byte +W(n): 14414799872.00 flops +T(n): 4864613888.00 cycles +P(n): 2.96 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 2250.17 flops/byte +W(n): 57629601792.00 flops +T(n): 20916678656.00 cycles +P(n): 2.76 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800369920.00 flops +Q(n) per iteration: 320056 bytes +I(n): 5625.17 flops/byte +W(n): 360073986048.00 flops +T(n): 154187153408.00 cycles +P(n): 2.34 flops/cycle + diff --git a/measurements/variant_float_inlining_SoA_symmetry/arm/1.log b/measurements/variant_float_inlining_SoA_symmetry/arm/1.log index 5cf9b82..0a29737 100644 --- a/measurements/variant_float_inlining_SoA_symmetry/arm/1.log +++ b/measurements/variant_float_inlining_SoA_symmetry/arm/1.log @@ -5,54 +5,54 @@ W(n) per iteration: 95500.00 flops Q(n) per iteration: 3256 bytes I(n): 29.33 flops/byte W(n): 19100000.00 flops -T(n): 12284355.00 cycles -P(n): 1.55 flops/cycle +T(n): 12254479.00 cycles +P(n): 1.56 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 371000.00 flops Q(n) per iteration: 6456 bytes I(n): 57.47 flops/byte W(n): 74200000.00 flops -T(n): 50450768.00 cycles -P(n): 1.47 flops/cycle +T(n): 50135688.00 cycles +P(n): 1.48 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 1462000.00 flops Q(n) per iteration: 12856 bytes I(n): 113.72 flops/byte W(n): 292400000.00 flops -T(n): 206580544.00 cycles -P(n): 1.42 flops/cycle +T(n): 205089712.00 cycles +P(n): 1.43 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 9055000.00 flops Q(n) per iteration: 32056 bytes I(n): 282.47 flops/byte W(n): 1811000064.00 flops -T(n): 1437377152.00 cycles -P(n): 1.26 flops/cycle +T(n): 1389706496.00 cycles +P(n): 1.30 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 36110000.00 flops Q(n) per iteration: 64056 bytes I(n): 563.73 flops/byte W(n): 7222000128.00 flops -T(n): 6982579200.00 cycles -P(n): 1.03 flops/cycle +T(n): 6575198208.00 cycles +P(n): 1.10 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 144220000.00 flops Q(n) per iteration: 128056 bytes I(n): 1126.23 flops/byte W(n): 28843999232.00 flops -T(n): 32683665408.00 cycles -P(n): 0.88 flops/cycle +T(n): 28691492864.00 cycles +P(n): 1.01 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 900550016.00 flops Q(n) per iteration: 320056 bytes I(n): 2813.73 flops/byte W(n): 180110000128.00 flops -T(n): 232878047232.00 cycles -P(n): 0.77 flops/cycle +T(n): 219295383552.00 cycles +P(n): 0.82 flops/cycle diff --git a/measurements/variant_float_inlining_SoA_symmetry/arm/2.log b/measurements/variant_float_inlining_SoA_symmetry/arm/2.log new file mode 100644 index 0000000..a4a39e3 --- /dev/null +++ b/measurements/variant_float_inlining_SoA_symmetry/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float_symmetry.cpp ./variants/variant_float_inlining_SoA_symmetry.cpp -o main_float_inlining_SoA_symmetry -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 3256 bytes +I(n): 29.33 flops/byte +W(n): 19100000.00 flops +T(n): 12256982.00 cycles +P(n): 1.56 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 6456 bytes +I(n): 57.47 flops/byte +W(n): 74200000.00 flops +T(n): 50311960.00 cycles +P(n): 1.47 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 12856 bytes +I(n): 113.72 flops/byte +W(n): 292400000.00 flops +T(n): 205116992.00 cycles +P(n): 1.43 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 282.47 flops/byte +W(n): 1811000064.00 flops +T(n): 1430680320.00 cycles +P(n): 1.27 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 563.73 flops/byte +W(n): 7222000128.00 flops +T(n): 6433722368.00 cycles +P(n): 1.12 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 1126.23 flops/byte +W(n): 28843999232.00 flops +T(n): 28671172608.00 cycles +P(n): 1.01 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550016.00 flops +Q(n) per iteration: 320056 bytes +I(n): 2813.73 flops/byte +W(n): 180110000128.00 flops +T(n): 226077196288.00 cycles +P(n): 0.80 flops/cycle + diff --git a/measurements/variant_float_inlining_SoA_symmetry/arm/3.log b/measurements/variant_float_inlining_SoA_symmetry/arm/3.log new file mode 100644 index 0000000..c3815be --- /dev/null +++ b/measurements/variant_float_inlining_SoA_symmetry/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float_symmetry.cpp ./variants/variant_float_inlining_SoA_symmetry.cpp -o main_float_inlining_SoA_symmetry -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 3256 bytes +I(n): 29.33 flops/byte +W(n): 19100000.00 flops +T(n): 12241891.00 cycles +P(n): 1.56 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 6456 bytes +I(n): 57.47 flops/byte +W(n): 74200000.00 flops +T(n): 50166764.00 cycles +P(n): 1.48 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 12856 bytes +I(n): 113.72 flops/byte +W(n): 292400000.00 flops +T(n): 205702832.00 cycles +P(n): 1.42 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 282.47 flops/byte +W(n): 1811000064.00 flops +T(n): 1447650560.00 cycles +P(n): 1.25 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 563.73 flops/byte +W(n): 7222000128.00 flops +T(n): 6788238336.00 cycles +P(n): 1.06 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 1126.23 flops/byte +W(n): 28843999232.00 flops +T(n): 28329756672.00 cycles +P(n): 1.02 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550016.00 flops +Q(n) per iteration: 320056 bytes +I(n): 2813.73 flops/byte +W(n): 180110000128.00 flops +T(n): 226648129536.00 cycles +P(n): 0.79 flops/cycle + diff --git a/measurements/variant_float_inlining_SoA_symmetry_ILP_16/arm/1.log b/measurements/variant_float_inlining_SoA_symmetry_ILP_16/arm/1.log index 4a9f516..f13fa58 100644 --- a/measurements/variant_float_inlining_SoA_symmetry_ILP_16/arm/1.log +++ b/measurements/variant_float_inlining_SoA_symmetry_ILP_16/arm/1.log @@ -5,47 +5,47 @@ W(n) per iteration: 95500.00 flops Q(n) per iteration: 3256 bytes I(n): 29.33 flops/byte W(n): 19100000.00 flops -T(n): 9097177.00 cycles -P(n): 2.10 flops/cycle +T(n): 9042754.00 cycles +P(n): 2.11 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 371000.00 flops Q(n) per iteration: 6456 bytes I(n): 57.47 flops/byte W(n): 74200000.00 flops -T(n): 33503178.00 cycles -P(n): 2.21 flops/cycle +T(n): 33334258.00 cycles +P(n): 2.23 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 1462000.00 flops Q(n) per iteration: 12856 bytes I(n): 113.72 flops/byte W(n): 292400000.00 flops -T(n): 118530416.00 cycles -P(n): 2.47 flops/cycle +T(n): 118737600.00 cycles +P(n): 2.46 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 9055000.00 flops Q(n) per iteration: 32056 bytes I(n): 282.47 flops/byte W(n): 1811000064.00 flops -T(n): 684755776.00 cycles -P(n): 2.64 flops/cycle +T(n): 683585216.00 cycles +P(n): 2.65 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 36110000.00 flops Q(n) per iteration: 64056 bytes I(n): 563.73 flops/byte W(n): 7222000128.00 flops -T(n): 2936483584.00 cycles -P(n): 2.46 flops/cycle +T(n): 2916843264.00 cycles +P(n): 2.48 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 144220000.00 flops Q(n) per iteration: 128056 bytes I(n): 1126.23 flops/byte W(n): 28843999232.00 flops -T(n): 12939599872.00 cycles +T(n): 12947107840.00 cycles P(n): 2.23 flops/cycle === N_P=5000 N_ITER=200 === @@ -53,6 +53,6 @@ W(n) per iteration: 900550016.00 flops Q(n) per iteration: 320056 bytes I(n): 2813.73 flops/byte W(n): 180110000128.00 flops -T(n): 156191883264.00 cycles -P(n): 1.15 flops/cycle +T(n): 154301726720.00 cycles +P(n): 1.17 flops/cycle diff --git a/measurements/variant_float_inlining_SoA_symmetry_ILP_16/arm/2.log b/measurements/variant_float_inlining_SoA_symmetry_ILP_16/arm/2.log new file mode 100644 index 0000000..994dbb1 --- /dev/null +++ b/measurements/variant_float_inlining_SoA_symmetry_ILP_16/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float_symmetry.cpp ./variants/variant_float_inlining_SoA_symmetry_ILP_16.cpp -o main_float_inlining_SoA_symmetry_ILP_16 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 3256 bytes +I(n): 29.33 flops/byte +W(n): 19100000.00 flops +T(n): 9068705.00 cycles +P(n): 2.11 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 6456 bytes +I(n): 57.47 flops/byte +W(n): 74200000.00 flops +T(n): 33619560.00 cycles +P(n): 2.21 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 12856 bytes +I(n): 113.72 flops/byte +W(n): 292400000.00 flops +T(n): 119648512.00 cycles +P(n): 2.44 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 282.47 flops/byte +W(n): 1811000064.00 flops +T(n): 688923456.00 cycles +P(n): 2.63 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 563.73 flops/byte +W(n): 7222000128.00 flops +T(n): 2882365184.00 cycles +P(n): 2.51 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 1126.23 flops/byte +W(n): 28843999232.00 flops +T(n): 12853307392.00 cycles +P(n): 2.24 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550016.00 flops +Q(n) per iteration: 320056 bytes +I(n): 2813.73 flops/byte +W(n): 180110000128.00 flops +T(n): 157621452800.00 cycles +P(n): 1.14 flops/cycle + diff --git a/measurements/variant_float_inlining_SoA_symmetry_ILP_16/arm/3.log b/measurements/variant_float_inlining_SoA_symmetry_ILP_16/arm/3.log new file mode 100644 index 0000000..10c963a --- /dev/null +++ b/measurements/variant_float_inlining_SoA_symmetry_ILP_16/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float_symmetry.cpp ./variants/variant_float_inlining_SoA_symmetry_ILP_16.cpp -o main_float_inlining_SoA_symmetry_ILP_16 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 3256 bytes +I(n): 29.33 flops/byte +W(n): 19100000.00 flops +T(n): 9045719.00 cycles +P(n): 2.11 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 6456 bytes +I(n): 57.47 flops/byte +W(n): 74200000.00 flops +T(n): 33367980.00 cycles +P(n): 2.22 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 12856 bytes +I(n): 113.72 flops/byte +W(n): 292400000.00 flops +T(n): 118491888.00 cycles +P(n): 2.47 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 282.47 flops/byte +W(n): 1811000064.00 flops +T(n): 683712000.00 cycles +P(n): 2.65 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 563.73 flops/byte +W(n): 7222000128.00 flops +T(n): 2850022656.00 cycles +P(n): 2.53 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 1126.23 flops/byte +W(n): 28843999232.00 flops +T(n): 12636773376.00 cycles +P(n): 2.28 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550016.00 flops +Q(n) per iteration: 320056 bytes +I(n): 2813.73 flops/byte +W(n): 180110000128.00 flops +T(n): 158985289728.00 cycles +P(n): 1.13 flops/cycle + diff --git a/measurements/variant_float_inlining_SoA_symmetry_ILP_2/arm/1.log b/measurements/variant_float_inlining_SoA_symmetry_ILP_2/arm/1.log index 44ae16b..698c123 100644 --- a/measurements/variant_float_inlining_SoA_symmetry_ILP_2/arm/1.log +++ b/measurements/variant_float_inlining_SoA_symmetry_ILP_2/arm/1.log @@ -5,54 +5,54 @@ W(n) per iteration: 95500.00 flops Q(n) per iteration: 3256 bytes I(n): 29.33 flops/byte W(n): 19100000.00 flops -T(n): 9170654.00 cycles -P(n): 2.08 flops/cycle +T(n): 9159459.00 cycles +P(n): 2.09 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 371000.00 flops Q(n) per iteration: 6456 bytes I(n): 57.47 flops/byte W(n): 74200000.00 flops -T(n): 33103448.00 cycles -P(n): 2.24 flops/cycle +T(n): 32975330.00 cycles +P(n): 2.25 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 1462000.00 flops Q(n) per iteration: 12856 bytes I(n): 113.72 flops/byte W(n): 292400000.00 flops -T(n): 123532256.00 cycles -P(n): 2.37 flops/cycle +T(n): 122404672.00 cycles +P(n): 2.39 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 9055000.00 flops Q(n) per iteration: 32056 bytes I(n): 282.47 flops/byte W(n): 1811000064.00 flops -T(n): 768825216.00 cycles -P(n): 2.36 flops/cycle +T(n): 770166592.00 cycles +P(n): 2.35 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 36110000.00 flops Q(n) per iteration: 64056 bytes I(n): 563.73 flops/byte W(n): 7222000128.00 flops -T(n): 3933374208.00 cycles -P(n): 1.84 flops/cycle +T(n): 3881816064.00 cycles +P(n): 1.86 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 144220000.00 flops Q(n) per iteration: 128056 bytes I(n): 1126.23 flops/byte W(n): 28843999232.00 flops -T(n): 18220894208.00 cycles -P(n): 1.58 flops/cycle +T(n): 17828823040.00 cycles +P(n): 1.62 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 900550016.00 flops Q(n) per iteration: 320056 bytes I(n): 2813.73 flops/byte W(n): 180110000128.00 flops -T(n): 154008551424.00 cycles -P(n): 1.17 flops/cycle +T(n): 151776935936.00 cycles +P(n): 1.19 flops/cycle diff --git a/measurements/variant_float_inlining_SoA_symmetry_ILP_2/arm/2.log b/measurements/variant_float_inlining_SoA_symmetry_ILP_2/arm/2.log new file mode 100644 index 0000000..11af37b --- /dev/null +++ b/measurements/variant_float_inlining_SoA_symmetry_ILP_2/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float_symmetry.cpp ./variants/variant_float_inlining_SoA_symmetry_ILP_2.cpp -o main_float_inlining_SoA_symmetry_ILP_2 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 3256 bytes +I(n): 29.33 flops/byte +W(n): 19100000.00 flops +T(n): 9193534.00 cycles +P(n): 2.08 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 6456 bytes +I(n): 57.47 flops/byte +W(n): 74200000.00 flops +T(n): 33038872.00 cycles +P(n): 2.25 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 12856 bytes +I(n): 113.72 flops/byte +W(n): 292400000.00 flops +T(n): 122987592.00 cycles +P(n): 2.38 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 282.47 flops/byte +W(n): 1811000064.00 flops +T(n): 770936704.00 cycles +P(n): 2.35 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 563.73 flops/byte +W(n): 7222000128.00 flops +T(n): 3915471104.00 cycles +P(n): 1.84 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 1126.23 flops/byte +W(n): 28843999232.00 flops +T(n): 18003859456.00 cycles +P(n): 1.60 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550016.00 flops +Q(n) per iteration: 320056 bytes +I(n): 2813.73 flops/byte +W(n): 180110000128.00 flops +T(n): 152343429120.00 cycles +P(n): 1.18 flops/cycle + diff --git a/measurements/variant_float_inlining_SoA_symmetry_ILP_2/arm/3.log b/measurements/variant_float_inlining_SoA_symmetry_ILP_2/arm/3.log new file mode 100644 index 0000000..958bfc5 --- /dev/null +++ b/measurements/variant_float_inlining_SoA_symmetry_ILP_2/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float_symmetry.cpp ./variants/variant_float_inlining_SoA_symmetry_ILP_2.cpp -o main_float_inlining_SoA_symmetry_ILP_2 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 3256 bytes +I(n): 29.33 flops/byte +W(n): 19100000.00 flops +T(n): 9172994.00 cycles +P(n): 2.08 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 6456 bytes +I(n): 57.47 flops/byte +W(n): 74200000.00 flops +T(n): 33106616.00 cycles +P(n): 2.24 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 12856 bytes +I(n): 113.72 flops/byte +W(n): 292400000.00 flops +T(n): 122288848.00 cycles +P(n): 2.39 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 282.47 flops/byte +W(n): 1811000064.00 flops +T(n): 771513856.00 cycles +P(n): 2.35 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 563.73 flops/byte +W(n): 7222000128.00 flops +T(n): 3915929344.00 cycles +P(n): 1.84 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 1126.23 flops/byte +W(n): 28843999232.00 flops +T(n): 18206220288.00 cycles +P(n): 1.58 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550016.00 flops +Q(n) per iteration: 320056 bytes +I(n): 2813.73 flops/byte +W(n): 180110000128.00 flops +T(n): 148517683200.00 cycles +P(n): 1.21 flops/cycle + diff --git a/measurements/variant_float_inlining_SoA_symmetry_ILP_4/arm/1.log b/measurements/variant_float_inlining_SoA_symmetry_ILP_4/arm/1.log index 02eb146..9da2feb 100644 --- a/measurements/variant_float_inlining_SoA_symmetry_ILP_4/arm/1.log +++ b/measurements/variant_float_inlining_SoA_symmetry_ILP_4/arm/1.log @@ -5,15 +5,15 @@ W(n) per iteration: 95500.00 flops Q(n) per iteration: 3256 bytes I(n): 29.33 flops/byte W(n): 19100000.00 flops -T(n): 9431654.00 cycles -P(n): 2.03 flops/cycle +T(n): 9376726.00 cycles +P(n): 2.04 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 371000.00 flops Q(n) per iteration: 6456 bytes I(n): 57.47 flops/byte W(n): 74200000.00 flops -T(n): 33563136.00 cycles +T(n): 33509446.00 cycles P(n): 2.21 flops/cycle === N_P=200 N_ITER=200 === @@ -21,23 +21,23 @@ W(n) per iteration: 1462000.00 flops Q(n) per iteration: 12856 bytes I(n): 113.72 flops/byte W(n): 292400000.00 flops -T(n): 118019376.00 cycles -P(n): 2.48 flops/cycle +T(n): 118757464.00 cycles +P(n): 2.46 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 9055000.00 flops Q(n) per iteration: 32056 bytes I(n): 282.47 flops/byte W(n): 1811000064.00 flops -T(n): 678422208.00 cycles -P(n): 2.67 flops/cycle +T(n): 676173376.00 cycles +P(n): 2.68 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 36110000.00 flops Q(n) per iteration: 64056 bytes I(n): 563.73 flops/byte W(n): 7222000128.00 flops -T(n): 2959266048.00 cycles +T(n): 2963177216.00 cycles P(n): 2.44 flops/cycle === N_P=2000 N_ITER=200 === @@ -45,7 +45,7 @@ W(n) per iteration: 144220000.00 flops Q(n) per iteration: 128056 bytes I(n): 1126.23 flops/byte W(n): 28843999232.00 flops -T(n): 12601305088.00 cycles +T(n): 12573572096.00 cycles P(n): 2.29 flops/cycle === N_P=5000 N_ITER=200 === @@ -53,6 +53,6 @@ W(n) per iteration: 900550016.00 flops Q(n) per iteration: 320056 bytes I(n): 2813.73 flops/byte W(n): 180110000128.00 flops -T(n): 144715956224.00 cycles -P(n): 1.24 flops/cycle +T(n): 143544270848.00 cycles +P(n): 1.25 flops/cycle diff --git a/measurements/variant_float_inlining_SoA_symmetry_ILP_4/arm/2.log b/measurements/variant_float_inlining_SoA_symmetry_ILP_4/arm/2.log new file mode 100644 index 0000000..073a1fb --- /dev/null +++ b/measurements/variant_float_inlining_SoA_symmetry_ILP_4/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float_symmetry.cpp ./variants/variant_float_inlining_SoA_symmetry_ILP_4.cpp -o main_float_inlining_SoA_symmetry_ILP_4 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 3256 bytes +I(n): 29.33 flops/byte +W(n): 19100000.00 flops +T(n): 9522074.00 cycles +P(n): 2.01 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 6456 bytes +I(n): 57.47 flops/byte +W(n): 74200000.00 flops +T(n): 33704588.00 cycles +P(n): 2.20 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 12856 bytes +I(n): 113.72 flops/byte +W(n): 292400000.00 flops +T(n): 119636704.00 cycles +P(n): 2.44 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 282.47 flops/byte +W(n): 1811000064.00 flops +T(n): 733238656.00 cycles +P(n): 2.47 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 563.73 flops/byte +W(n): 7222000128.00 flops +T(n): 3048842752.00 cycles +P(n): 2.37 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 1126.23 flops/byte +W(n): 28843999232.00 flops +T(n): 13217501184.00 cycles +P(n): 2.18 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550016.00 flops +Q(n) per iteration: 320056 bytes +I(n): 2813.73 flops/byte +W(n): 180110000128.00 flops +T(n): 145096245248.00 cycles +P(n): 1.24 flops/cycle + diff --git a/measurements/variant_float_inlining_SoA_symmetry_ILP_4/arm/3.log b/measurements/variant_float_inlining_SoA_symmetry_ILP_4/arm/3.log new file mode 100644 index 0000000..f24cbf7 --- /dev/null +++ b/measurements/variant_float_inlining_SoA_symmetry_ILP_4/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float_symmetry.cpp ./variants/variant_float_inlining_SoA_symmetry_ILP_4.cpp -o main_float_inlining_SoA_symmetry_ILP_4 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 3256 bytes +I(n): 29.33 flops/byte +W(n): 19100000.00 flops +T(n): 9386540.00 cycles +P(n): 2.03 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 6456 bytes +I(n): 57.47 flops/byte +W(n): 74200000.00 flops +T(n): 33523044.00 cycles +P(n): 2.21 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 12856 bytes +I(n): 113.72 flops/byte +W(n): 292400000.00 flops +T(n): 119032680.00 cycles +P(n): 2.46 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 282.47 flops/byte +W(n): 1811000064.00 flops +T(n): 676348160.00 cycles +P(n): 2.68 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 563.73 flops/byte +W(n): 7222000128.00 flops +T(n): 2969299712.00 cycles +P(n): 2.43 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 1126.23 flops/byte +W(n): 28843999232.00 flops +T(n): 12426423296.00 cycles +P(n): 2.32 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550016.00 flops +Q(n) per iteration: 320056 bytes +I(n): 2813.73 flops/byte +W(n): 180110000128.00 flops +T(n): 140797083648.00 cycles +P(n): 1.28 flops/cycle + diff --git a/measurements/variant_float_inlining_SoA_symmetry_ILP_8/arm/1.log b/measurements/variant_float_inlining_SoA_symmetry_ILP_8/arm/1.log index ea3e04a..df9ff9c 100644 --- a/measurements/variant_float_inlining_SoA_symmetry_ILP_8/arm/1.log +++ b/measurements/variant_float_inlining_SoA_symmetry_ILP_8/arm/1.log @@ -5,15 +5,15 @@ W(n) per iteration: 95500.00 flops Q(n) per iteration: 3256 bytes I(n): 29.33 flops/byte W(n): 19100000.00 flops -T(n): 8858727.00 cycles -P(n): 2.16 flops/cycle +T(n): 8798921.00 cycles +P(n): 2.17 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 371000.00 flops Q(n) per iteration: 6456 bytes I(n): 57.47 flops/byte W(n): 74200000.00 flops -T(n): 35191192.00 cycles +T(n): 35175424.00 cycles P(n): 2.11 flops/cycle === N_P=200 N_ITER=200 === @@ -21,38 +21,38 @@ W(n) per iteration: 1462000.00 flops Q(n) per iteration: 12856 bytes I(n): 113.72 flops/byte W(n): 292400000.00 flops -T(n): 136348512.00 cycles -P(n): 2.14 flops/cycle +T(n): 135894384.00 cycles +P(n): 2.15 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 9055000.00 flops Q(n) per iteration: 32056 bytes I(n): 282.47 flops/byte W(n): 1811000064.00 flops -T(n): 881212544.00 cycles -P(n): 2.06 flops/cycle +T(n): 889473408.00 cycles +P(n): 2.04 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 36110000.00 flops Q(n) per iteration: 64056 bytes I(n): 563.73 flops/byte W(n): 7222000128.00 flops -T(n): 4358949376.00 cycles -P(n): 1.66 flops/cycle +T(n): 4264077312.00 cycles +P(n): 1.69 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 144220000.00 flops Q(n) per iteration: 128056 bytes I(n): 1126.23 flops/byte W(n): 28843999232.00 flops -T(n): 19275378688.00 cycles -P(n): 1.50 flops/cycle +T(n): 19037489152.00 cycles +P(n): 1.52 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 900550016.00 flops Q(n) per iteration: 320056 bytes I(n): 2813.73 flops/byte W(n): 180110000128.00 flops -T(n): 158142201856.00 cycles -P(n): 1.14 flops/cycle +T(n): 156173221888.00 cycles +P(n): 1.15 flops/cycle diff --git a/measurements/variant_float_inlining_SoA_symmetry_ILP_8/arm/2.log b/measurements/variant_float_inlining_SoA_symmetry_ILP_8/arm/2.log new file mode 100644 index 0000000..c12f33c --- /dev/null +++ b/measurements/variant_float_inlining_SoA_symmetry_ILP_8/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float_symmetry.cpp ./variants/variant_float_inlining_SoA_symmetry_ILP_8.cpp -o main_float_inlining_SoA_symmetry_ILP_8 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 3256 bytes +I(n): 29.33 flops/byte +W(n): 19100000.00 flops +T(n): 8862931.00 cycles +P(n): 2.16 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 6456 bytes +I(n): 57.47 flops/byte +W(n): 74200000.00 flops +T(n): 35686564.00 cycles +P(n): 2.08 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 12856 bytes +I(n): 113.72 flops/byte +W(n): 292400000.00 flops +T(n): 137157216.00 cycles +P(n): 2.13 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 282.47 flops/byte +W(n): 1811000064.00 flops +T(n): 887476736.00 cycles +P(n): 2.04 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 563.73 flops/byte +W(n): 7222000128.00 flops +T(n): 4329847808.00 cycles +P(n): 1.67 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 1126.23 flops/byte +W(n): 28843999232.00 flops +T(n): 19318816768.00 cycles +P(n): 1.49 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550016.00 flops +Q(n) per iteration: 320056 bytes +I(n): 2813.73 flops/byte +W(n): 180110000128.00 flops +T(n): 157832298496.00 cycles +P(n): 1.14 flops/cycle + diff --git a/measurements/variant_float_inlining_SoA_symmetry_ILP_8/arm/3.log b/measurements/variant_float_inlining_SoA_symmetry_ILP_8/arm/3.log new file mode 100644 index 0000000..1c02ce4 --- /dev/null +++ b/measurements/variant_float_inlining_SoA_symmetry_ILP_8/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-float_symmetry.cpp ./variants/variant_float_inlining_SoA_symmetry_ILP_8.cpp -o main_float_inlining_SoA_symmetry_ILP_8 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 3256 bytes +I(n): 29.33 flops/byte +W(n): 19100000.00 flops +T(n): 8850936.00 cycles +P(n): 2.16 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 6456 bytes +I(n): 57.47 flops/byte +W(n): 74200000.00 flops +T(n): 35237568.00 cycles +P(n): 2.11 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 12856 bytes +I(n): 113.72 flops/byte +W(n): 292400000.00 flops +T(n): 137259488.00 cycles +P(n): 2.13 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 32056 bytes +I(n): 282.47 flops/byte +W(n): 1811000064.00 flops +T(n): 894708160.00 cycles +P(n): 2.02 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 64056 bytes +I(n): 563.73 flops/byte +W(n): 7222000128.00 flops +T(n): 4392257536.00 cycles +P(n): 1.64 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 128056 bytes +I(n): 1126.23 flops/byte +W(n): 28843999232.00 flops +T(n): 19428847616.00 cycles +P(n): 1.48 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550016.00 flops +Q(n) per iteration: 320056 bytes +I(n): 2813.73 flops/byte +W(n): 180110000128.00 flops +T(n): 157138927616.00 cycles +P(n): 1.15 flops/cycle + diff --git a/measurements/variant_inlining_7/arm/1.log b/measurements/variant_inlining_7/arm/1.log index f4d24ee..f4ac518 100644 --- a/measurements/variant_inlining_7/arm/1.log +++ b/measurements/variant_inlining_7/arm/1.log @@ -5,39 +5,39 @@ W(n) per iteration: 183700.00 flops Q(n) per iteration: 6512 bytes I(n): 28.21 flops/byte W(n): 36740000.00 flops -T(n): 16465719.25 cycles -P(n): 2.23 flops/cycle +T(n): 16982433.12 cycles +P(n): 2.16 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 727400.00 flops Q(n) per iteration: 12912 bytes I(n): 56.34 flops/byte W(n): 145480000.00 flops -T(n): 68986819.50 cycles -P(n): 2.11 flops/cycle +T(n): 69901888.50 cycles +P(n): 2.08 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 2894800.00 flops Q(n) per iteration: 25712 bytes I(n): 112.59 flops/byte W(n): 578960000.00 flops -T(n): 296100633.00 cycles -P(n): 1.96 flops/cycle +T(n): 302888308.00 cycles +P(n): 1.91 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 18037000.00 flops Q(n) per iteration: 64112 bytes I(n): 281.34 flops/byte W(n): 3607400000.00 flops -T(n): 2015322383.00 cycles -P(n): 1.79 flops/cycle +T(n): 2052868152.00 cycles +P(n): 1.76 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 72074000.00 flops Q(n) per iteration: 128112 bytes I(n): 562.59 flops/byte W(n): 14414800000.00 flops -T(n): 8425401169.00 cycles +T(n): 8417635258.00 cycles P(n): 1.71 flops/cycle === N_P=2000 N_ITER=200 === @@ -45,14 +45,14 @@ W(n) per iteration: 288148000.00 flops Q(n) per iteration: 256112 bytes I(n): 1125.09 flops/byte W(n): 57629600000.00 flops -T(n): 35087107382.00 cycles -P(n): 1.64 flops/cycle +T(n): 35363573917.00 cycles +P(n): 1.63 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 1800370000.00 flops Q(n) per iteration: 640112 bytes I(n): 2812.59 flops/byte W(n): 360074000000.00 flops -T(n): 225285744034.00 cycles -P(n): 1.60 flops/cycle +T(n): 228532646195.00 cycles +P(n): 1.58 flops/cycle diff --git a/measurements/variant_inlining_7/arm/2.log b/measurements/variant_inlining_7/arm/2.log new file mode 100644 index 0000000..6f3500b --- /dev/null +++ b/measurements/variant_inlining_7/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_inlining_7.cpp -o main_inlining_7 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 16532863.38 cycles +P(n): 2.22 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 68951749.50 cycles +P(n): 2.11 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 297964775.00 cycles +P(n): 1.94 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 2043353865.00 cycles +P(n): 1.77 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 8411006888.00 cycles +P(n): 1.71 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 35078318181.00 cycles +P(n): 1.64 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 226390920838.00 cycles +P(n): 1.59 flops/cycle + diff --git a/measurements/variant_inlining_7/arm/3.log b/measurements/variant_inlining_7/arm/3.log new file mode 100644 index 0000000..5cdf5ff --- /dev/null +++ b/measurements/variant_inlining_7/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_inlining_7.cpp -o main_inlining_7 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 16539475.50 cycles +P(n): 2.22 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 68733742.50 cycles +P(n): 2.12 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 299082818.00 cycles +P(n): 1.94 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 2023139542.00 cycles +P(n): 1.78 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 8391580635.00 cycles +P(n): 1.72 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 35018999759.00 cycles +P(n): 1.65 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 225594949413.00 cycles +P(n): 1.60 flops/cycle + diff --git a/measurements/variant_inlining_7_SoA/arm/1.log b/measurements/variant_inlining_7_SoA/arm/1.log index 4b05723..25787e6 100644 --- a/measurements/variant_inlining_7_SoA/arm/1.log +++ b/measurements/variant_inlining_7_SoA/arm/1.log @@ -5,23 +5,23 @@ W(n) per iteration: 183700.00 flops Q(n) per iteration: 6512 bytes I(n): 28.21 flops/byte W(n): 36740000.00 flops -T(n): 15734093.62 cycles -P(n): 2.34 flops/cycle +T(n): 15742093.00 cycles +P(n): 2.33 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 727400.00 flops Q(n) per iteration: 12912 bytes I(n): 56.34 flops/byte W(n): 145480000.00 flops -T(n): 63384057.50 cycles -P(n): 2.30 flops/cycle +T(n): 62047064.00 cycles +P(n): 2.34 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 2894800.00 flops Q(n) per iteration: 25712 bytes I(n): 112.59 flops/byte W(n): 578960000.00 flops -T(n): 258485217.00 cycles +T(n): 258760280.00 cycles P(n): 2.24 flops/cycle === N_P=500 N_ITER=200 === @@ -29,7 +29,7 @@ W(n) per iteration: 18037000.00 flops Q(n) per iteration: 64112 bytes I(n): 281.34 flops/byte W(n): 3607400000.00 flops -T(n): 1806293743.00 cycles +T(n): 1801642548.00 cycles P(n): 2.00 flops/cycle === N_P=1000 N_ITER=200 === @@ -37,7 +37,7 @@ W(n) per iteration: 72074000.00 flops Q(n) per iteration: 128112 bytes I(n): 562.59 flops/byte W(n): 14414800000.00 flops -T(n): 7331209997.00 cycles +T(n): 7309015152.00 cycles P(n): 1.97 flops/cycle === N_P=2000 N_ITER=200 === @@ -45,7 +45,7 @@ W(n) per iteration: 288148000.00 flops Q(n) per iteration: 256112 bytes I(n): 1125.09 flops/byte W(n): 57629600000.00 flops -T(n): 31741034397.00 cycles +T(n): 31580795820.00 cycles P(n): 1.82 flops/cycle === N_P=5000 N_ITER=200 === @@ -53,6 +53,6 @@ W(n) per iteration: 1800370000.00 flops Q(n) per iteration: 640112 bytes I(n): 2812.59 flops/byte W(n): 360074000000.00 flops -T(n): 228332189160.00 cycles -P(n): 1.58 flops/cycle +T(n): 225252090306.00 cycles +P(n): 1.60 flops/cycle diff --git a/measurements/variant_inlining_7_SoA/arm/2.log b/measurements/variant_inlining_7_SoA/arm/2.log new file mode 100644 index 0000000..089a736 --- /dev/null +++ b/measurements/variant_inlining_7_SoA/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_inlining_7_SoA.cpp -o main_inlining_7_SoA -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 15726624.62 cycles +P(n): 2.34 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 62174486.50 cycles +P(n): 2.34 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 259603836.00 cycles +P(n): 2.23 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 1804324791.00 cycles +P(n): 2.00 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 7328087365.00 cycles +P(n): 1.97 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 31730627302.00 cycles +P(n): 1.82 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 227943175380.00 cycles +P(n): 1.58 flops/cycle + diff --git a/measurements/variant_inlining_7_SoA/arm/3.log b/measurements/variant_inlining_7_SoA/arm/3.log new file mode 100644 index 0000000..c432e14 --- /dev/null +++ b/measurements/variant_inlining_7_SoA/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_inlining_7_SoA.cpp -o main_inlining_7_SoA -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 15774522.50 cycles +P(n): 2.33 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 62462632.00 cycles +P(n): 2.33 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 259444271.00 cycles +P(n): 2.23 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 1810302610.00 cycles +P(n): 1.99 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 7364211387.00 cycles +P(n): 1.96 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 31976299325.00 cycles +P(n): 1.80 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 227745231274.00 cycles +P(n): 1.58 flops/cycle + diff --git a/measurements/variant_inlining_7_symmetry_1/arm/1.log b/measurements/variant_inlining_7_symmetry_1/arm/1.log index d49f009..05eabb6 100644 --- a/measurements/variant_inlining_7_symmetry_1/arm/1.log +++ b/measurements/variant_inlining_7_symmetry_1/arm/1.log @@ -5,23 +5,23 @@ W(n) per iteration: 95500.00 flops Q(n) per iteration: 6512 bytes I(n): 14.67 flops/byte W(n): 19100000.00 flops -T(n): 16259793.62 cycles -P(n): 1.17 flops/cycle +T(n): 16117516.88 cycles +P(n): 1.19 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 371000.00 flops Q(n) per iteration: 12912 bytes I(n): 28.73 flops/byte W(n): 74200000.00 flops -T(n): 65726236.50 cycles -P(n): 1.13 flops/cycle +T(n): 66078411.00 cycles +P(n): 1.12 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 1462000.00 flops Q(n) per iteration: 25712 bytes I(n): 56.86 flops/byte W(n): 292400000.00 flops -T(n): 279432230.00 cycles +T(n): 277193381.00 cycles P(n): 1.05 flops/cycle === N_P=500 N_ITER=200 === @@ -29,30 +29,30 @@ W(n) per iteration: 9055000.00 flops Q(n) per iteration: 64112 bytes I(n): 141.24 flops/byte W(n): 1811000000.00 flops -T(n): 2173883627.00 cycles -P(n): 0.83 flops/cycle +T(n): 1988191984.00 cycles +P(n): 0.91 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 36110000.00 flops Q(n) per iteration: 128112 bytes I(n): 281.86 flops/byte W(n): 7222000000.00 flops -T(n): 9352805794.00 cycles -P(n): 0.77 flops/cycle +T(n): 8569511226.00 cycles +P(n): 0.84 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 144220000.00 flops Q(n) per iteration: 256112 bytes I(n): 563.11 flops/byte W(n): 28844000000.00 flops -T(n): 41809462783.00 cycles -P(n): 0.69 flops/cycle +T(n): 41302583719.00 cycles +P(n): 0.70 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 900550000.00 flops Q(n) per iteration: 640112 bytes I(n): 1406.86 flops/byte W(n): 180110000000.00 flops -T(n): 284419980420.00 cycles -P(n): 0.63 flops/cycle +T(n): 279313154848.00 cycles +P(n): 0.64 flops/cycle diff --git a/measurements/variant_inlining_7_symmetry_1/arm/2.log b/measurements/variant_inlining_7_symmetry_1/arm/2.log new file mode 100644 index 0000000..f619d5f --- /dev/null +++ b/measurements/variant_inlining_7_symmetry_1/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-symmetry.cpp ./variants/variant_inlining_7_symmetry_1.cpp -o main_inlining_7_symmetry_1 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 6512 bytes +I(n): 14.67 flops/byte +W(n): 19100000.00 flops +T(n): 16282625.38 cycles +P(n): 1.17 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 12912 bytes +I(n): 28.73 flops/byte +W(n): 74200000.00 flops +T(n): 66366868.00 cycles +P(n): 1.12 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 25712 bytes +I(n): 56.86 flops/byte +W(n): 292400000.00 flops +T(n): 281595120.00 cycles +P(n): 1.04 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 141.24 flops/byte +W(n): 1811000000.00 flops +T(n): 2180955081.00 cycles +P(n): 0.83 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 281.86 flops/byte +W(n): 7222000000.00 flops +T(n): 9224037046.00 cycles +P(n): 0.78 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 563.11 flops/byte +W(n): 28844000000.00 flops +T(n): 40458642557.00 cycles +P(n): 0.71 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 1406.86 flops/byte +W(n): 180110000000.00 flops +T(n): 283134924244.00 cycles +P(n): 0.64 flops/cycle + diff --git a/measurements/variant_inlining_7_symmetry_1/arm/3.log b/measurements/variant_inlining_7_symmetry_1/arm/3.log new file mode 100644 index 0000000..ec1c8ea --- /dev/null +++ b/measurements/variant_inlining_7_symmetry_1/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-symmetry.cpp ./variants/variant_inlining_7_symmetry_1.cpp -o main_inlining_7_symmetry_1 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 6512 bytes +I(n): 14.67 flops/byte +W(n): 19100000.00 flops +T(n): 16211911.12 cycles +P(n): 1.18 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 12912 bytes +I(n): 28.73 flops/byte +W(n): 74200000.00 flops +T(n): 68639197.00 cycles +P(n): 1.08 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 25712 bytes +I(n): 56.86 flops/byte +W(n): 292400000.00 flops +T(n): 286606132.00 cycles +P(n): 1.02 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 141.24 flops/byte +W(n): 1811000000.00 flops +T(n): 2147698763.00 cycles +P(n): 0.84 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 281.86 flops/byte +W(n): 7222000000.00 flops +T(n): 9373351571.00 cycles +P(n): 0.77 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 563.11 flops/byte +W(n): 28844000000.00 flops +T(n): 41589198731.00 cycles +P(n): 0.69 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 1406.86 flops/byte +W(n): 180110000000.00 flops +T(n): 282863250817.00 cycles +P(n): 0.64 flops/cycle + diff --git a/measurements/variant_inlining_7_symmetry_2/arm/1.log b/measurements/variant_inlining_7_symmetry_2/arm/1.log index 3112a0e..1bdcc45 100644 --- a/measurements/variant_inlining_7_symmetry_2/arm/1.log +++ b/measurements/variant_inlining_7_symmetry_2/arm/1.log @@ -5,7 +5,7 @@ W(n) per iteration: 95500.00 flops Q(n) per iteration: 6512 bytes I(n): 14.67 flops/byte W(n): 19100000.00 flops -T(n): 16243406.75 cycles +T(n): 16142806.62 cycles P(n): 1.18 flops/cycle === N_P=100 N_ITER=200 === @@ -13,7 +13,7 @@ W(n) per iteration: 371000.00 flops Q(n) per iteration: 12912 bytes I(n): 28.73 flops/byte W(n): 74200000.00 flops -T(n): 66097465.00 cycles +T(n): 66409907.00 cycles P(n): 1.12 flops/cycle === N_P=200 N_ITER=200 === @@ -21,38 +21,38 @@ W(n) per iteration: 1462000.00 flops Q(n) per iteration: 25712 bytes I(n): 56.86 flops/byte W(n): 292400000.00 flops -T(n): 285031402.00 cycles -P(n): 1.03 flops/cycle +T(n): 278127300.00 cycles +P(n): 1.05 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 9055000.00 flops Q(n) per iteration: 64112 bytes I(n): 141.24 flops/byte W(n): 1811000000.00 flops -T(n): 2157696662.00 cycles -P(n): 0.84 flops/cycle +T(n): 1983862454.00 cycles +P(n): 0.91 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 36110000.00 flops Q(n) per iteration: 128112 bytes I(n): 281.86 flops/byte W(n): 7222000000.00 flops -T(n): 9630862508.00 cycles -P(n): 0.75 flops/cycle +T(n): 8764024847.00 cycles +P(n): 0.82 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 144220000.00 flops Q(n) per iteration: 256112 bytes I(n): 563.11 flops/byte W(n): 28844000000.00 flops -T(n): 41773155034.00 cycles -P(n): 0.69 flops/cycle +T(n): 39723210337.00 cycles +P(n): 0.73 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 900550000.00 flops Q(n) per iteration: 640112 bytes I(n): 1406.86 flops/byte W(n): 180110000000.00 flops -T(n): 276672447552.00 cycles +T(n): 277797538652.00 cycles P(n): 0.65 flops/cycle diff --git a/measurements/variant_inlining_7_symmetry_2/arm/2.log b/measurements/variant_inlining_7_symmetry_2/arm/2.log new file mode 100644 index 0000000..fbf876f --- /dev/null +++ b/measurements/variant_inlining_7_symmetry_2/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-symmetry.cpp ./variants/variant_inlining_7_symmetry_2.cpp -o main_inlining_7_symmetry_2 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 6512 bytes +I(n): 14.67 flops/byte +W(n): 19100000.00 flops +T(n): 16168869.00 cycles +P(n): 1.18 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 12912 bytes +I(n): 28.73 flops/byte +W(n): 74200000.00 flops +T(n): 65986593.50 cycles +P(n): 1.12 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 25712 bytes +I(n): 56.86 flops/byte +W(n): 292400000.00 flops +T(n): 276843557.00 cycles +P(n): 1.06 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 141.24 flops/byte +W(n): 1811000000.00 flops +T(n): 1995847484.00 cycles +P(n): 0.91 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 281.86 flops/byte +W(n): 7222000000.00 flops +T(n): 8724469474.00 cycles +P(n): 0.83 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 563.11 flops/byte +W(n): 28844000000.00 flops +T(n): 40376577856.00 cycles +P(n): 0.71 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 1406.86 flops/byte +W(n): 180110000000.00 flops +T(n): 277173820474.00 cycles +P(n): 0.65 flops/cycle + diff --git a/measurements/variant_inlining_7_symmetry_2/arm/3.log b/measurements/variant_inlining_7_symmetry_2/arm/3.log new file mode 100644 index 0000000..f137b32 --- /dev/null +++ b/measurements/variant_inlining_7_symmetry_2/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-symmetry.cpp ./variants/variant_inlining_7_symmetry_2.cpp -o main_inlining_7_symmetry_2 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 6512 bytes +I(n): 14.67 flops/byte +W(n): 19100000.00 flops +T(n): 16206126.50 cycles +P(n): 1.18 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 12912 bytes +I(n): 28.73 flops/byte +W(n): 74200000.00 flops +T(n): 65943577.00 cycles +P(n): 1.13 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 25712 bytes +I(n): 56.86 flops/byte +W(n): 292400000.00 flops +T(n): 278774560.00 cycles +P(n): 1.05 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 141.24 flops/byte +W(n): 1811000000.00 flops +T(n): 1997709210.00 cycles +P(n): 0.91 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 281.86 flops/byte +W(n): 7222000000.00 flops +T(n): 9295948552.00 cycles +P(n): 0.78 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 563.11 flops/byte +W(n): 28844000000.00 flops +T(n): 39953243724.00 cycles +P(n): 0.72 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 1406.86 flops/byte +W(n): 180110000000.00 flops +T(n): 272359437585.00 cycles +P(n): 0.66 flops/cycle + diff --git a/measurements/variant_inlining_7_symmetry_2_ilp_1/arm/1.log b/measurements/variant_inlining_7_symmetry_2_ilp_1/arm/1.log index 80d26c7..01a70ff 100644 --- a/measurements/variant_inlining_7_symmetry_2_ilp_1/arm/1.log +++ b/measurements/variant_inlining_7_symmetry_2_ilp_1/arm/1.log @@ -5,7 +5,7 @@ W(n) per iteration: 95500.00 flops Q(n) per iteration: 6512 bytes I(n): 14.67 flops/byte W(n): 19100000.00 flops -T(n): 16213663.88 cycles +T(n): 16143845.50 cycles P(n): 1.18 flops/cycle === N_P=100 N_ITER=200 === @@ -13,7 +13,7 @@ W(n) per iteration: 371000.00 flops Q(n) per iteration: 12912 bytes I(n): 28.73 flops/byte W(n): 74200000.00 flops -T(n): 64706500.00 cycles +T(n): 64419141.00 cycles P(n): 1.15 flops/cycle === N_P=200 N_ITER=200 === @@ -21,38 +21,38 @@ W(n) per iteration: 1462000.00 flops Q(n) per iteration: 25712 bytes I(n): 56.86 flops/byte W(n): 292400000.00 flops -T(n): 266479325.00 cycles -P(n): 1.10 flops/cycle +T(n): 262269150.00 cycles +P(n): 1.11 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 9055000.00 flops Q(n) per iteration: 64112 bytes I(n): 141.24 flops/byte W(n): 1811000000.00 flops -T(n): 1860823343.00 cycles -P(n): 0.97 flops/cycle +T(n): 1824103468.00 cycles +P(n): 0.99 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 36110000.00 flops Q(n) per iteration: 128112 bytes I(n): 281.86 flops/byte W(n): 7222000000.00 flops -T(n): 8047713694.00 cycles -P(n): 0.90 flops/cycle +T(n): 7918236003.00 cycles +P(n): 0.91 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 144220000.00 flops Q(n) per iteration: 256112 bytes I(n): 563.11 flops/byte W(n): 28844000000.00 flops -T(n): 35415664148.00 cycles -P(n): 0.81 flops/cycle +T(n): 34961836207.00 cycles +P(n): 0.83 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 900550000.00 flops Q(n) per iteration: 640112 bytes I(n): 1406.86 flops/byte W(n): 180110000000.00 flops -T(n): 188828167800.00 cycles -P(n): 0.95 flops/cycle +T(n): 187037217832.00 cycles +P(n): 0.96 flops/cycle diff --git a/measurements/variant_inlining_7_symmetry_2_ilp_1/arm/2.log b/measurements/variant_inlining_7_symmetry_2_ilp_1/arm/2.log new file mode 100644 index 0000000..56ab052 --- /dev/null +++ b/measurements/variant_inlining_7_symmetry_2_ilp_1/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-symmetry.cpp ./variants/variant_inlining_7_symmetry_2_ilp_1.cpp -o main_inlining_7_symmetry_2_ilp_1 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 6512 bytes +I(n): 14.67 flops/byte +W(n): 19100000.00 flops +T(n): 16226613.25 cycles +P(n): 1.18 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 12912 bytes +I(n): 28.73 flops/byte +W(n): 74200000.00 flops +T(n): 64525677.00 cycles +P(n): 1.15 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 25712 bytes +I(n): 56.86 flops/byte +W(n): 292400000.00 flops +T(n): 265742655.00 cycles +P(n): 1.10 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 141.24 flops/byte +W(n): 1811000000.00 flops +T(n): 1838249013.00 cycles +P(n): 0.99 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 281.86 flops/byte +W(n): 7222000000.00 flops +T(n): 8064528015.00 cycles +P(n): 0.90 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 563.11 flops/byte +W(n): 28844000000.00 flops +T(n): 35467168991.00 cycles +P(n): 0.81 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 1406.86 flops/byte +W(n): 180110000000.00 flops +T(n): 188195572787.00 cycles +P(n): 0.96 flops/cycle + diff --git a/measurements/variant_inlining_7_symmetry_2_ilp_1/arm/3.log b/measurements/variant_inlining_7_symmetry_2_ilp_1/arm/3.log new file mode 100644 index 0000000..4a36a35 --- /dev/null +++ b/measurements/variant_inlining_7_symmetry_2_ilp_1/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-symmetry.cpp ./variants/variant_inlining_7_symmetry_2_ilp_1.cpp -o main_inlining_7_symmetry_2_ilp_1 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 6512 bytes +I(n): 14.67 flops/byte +W(n): 19100000.00 flops +T(n): 16192093.00 cycles +P(n): 1.18 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 12912 bytes +I(n): 28.73 flops/byte +W(n): 74200000.00 flops +T(n): 64696094.50 cycles +P(n): 1.15 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 25712 bytes +I(n): 56.86 flops/byte +W(n): 292400000.00 flops +T(n): 262498790.00 cycles +P(n): 1.11 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 141.24 flops/byte +W(n): 1811000000.00 flops +T(n): 1847962807.00 cycles +P(n): 0.98 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 281.86 flops/byte +W(n): 7222000000.00 flops +T(n): 8023346263.00 cycles +P(n): 0.90 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 563.11 flops/byte +W(n): 28844000000.00 flops +T(n): 35945565456.00 cycles +P(n): 0.80 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 1406.86 flops/byte +W(n): 180110000000.00 flops +T(n): 188284363447.00 cycles +P(n): 0.96 flops/cycle + diff --git a/measurements/variant_inlining_7_symmetry_2_ilp_2/arm/1.log b/measurements/variant_inlining_7_symmetry_2_ilp_2/arm/1.log index 8a37901..8f59de4 100644 --- a/measurements/variant_inlining_7_symmetry_2_ilp_2/arm/1.log +++ b/measurements/variant_inlining_7_symmetry_2_ilp_2/arm/1.log @@ -5,15 +5,15 @@ W(n) per iteration: 95500.00 flops Q(n) per iteration: 6512 bytes I(n): 14.67 flops/byte W(n): 19100000.00 flops -T(n): 16951060.50 cycles -P(n): 1.13 flops/cycle +T(n): 16580100.12 cycles +P(n): 1.15 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 371000.00 flops Q(n) per iteration: 12912 bytes I(n): 28.73 flops/byte W(n): 74200000.00 flops -T(n): 65369333.50 cycles +T(n): 65024621.00 cycles P(n): 1.14 flops/cycle === N_P=200 N_ITER=200 === @@ -21,38 +21,38 @@ W(n) per iteration: 1462000.00 flops Q(n) per iteration: 25712 bytes I(n): 56.86 flops/byte W(n): 292400000.00 flops -T(n): 264705723.00 cycles -P(n): 1.10 flops/cycle +T(n): 261750350.00 cycles +P(n): 1.12 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 9055000.00 flops Q(n) per iteration: 64112 bytes I(n): 141.24 flops/byte W(n): 1811000000.00 flops -T(n): 1782045217.00 cycles -P(n): 1.02 flops/cycle +T(n): 1758395916.00 cycles +P(n): 1.03 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 36110000.00 flops Q(n) per iteration: 128112 bytes I(n): 281.86 flops/byte W(n): 7222000000.00 flops -T(n): 7925808103.00 cycles -P(n): 0.91 flops/cycle +T(n): 7434680291.00 cycles +P(n): 0.97 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 144220000.00 flops Q(n) per iteration: 256112 bytes I(n): 563.11 flops/byte W(n): 28844000000.00 flops -T(n): 35824664520.00 cycles -P(n): 0.81 flops/cycle +T(n): 34379672055.00 cycles +P(n): 0.84 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 900550000.00 flops Q(n) per iteration: 640112 bytes I(n): 1406.86 flops/byte W(n): 180110000000.00 flops -T(n): 247389191708.00 cycles -P(n): 0.73 flops/cycle +T(n): 239178588140.00 cycles +P(n): 0.75 flops/cycle diff --git a/measurements/variant_inlining_7_symmetry_2_ilp_2/arm/2.log b/measurements/variant_inlining_7_symmetry_2_ilp_2/arm/2.log new file mode 100644 index 0000000..304b068 --- /dev/null +++ b/measurements/variant_inlining_7_symmetry_2_ilp_2/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-symmetry.cpp ./variants/variant_inlining_7_symmetry_2_ilp_2.cpp -o main_inlining_7_symmetry_2_ilp_2 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 6512 bytes +I(n): 14.67 flops/byte +W(n): 19100000.00 flops +T(n): 16547697.88 cycles +P(n): 1.15 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 12912 bytes +I(n): 28.73 flops/byte +W(n): 74200000.00 flops +T(n): 65340354.00 cycles +P(n): 1.14 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 25712 bytes +I(n): 56.86 flops/byte +W(n): 292400000.00 flops +T(n): 262140162.00 cycles +P(n): 1.12 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 141.24 flops/byte +W(n): 1811000000.00 flops +T(n): 1791946939.00 cycles +P(n): 1.01 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 281.86 flops/byte +W(n): 7222000000.00 flops +T(n): 7559703643.00 cycles +P(n): 0.96 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 563.11 flops/byte +W(n): 28844000000.00 flops +T(n): 34850899212.00 cycles +P(n): 0.83 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 1406.86 flops/byte +W(n): 180110000000.00 flops +T(n): 239389045900.00 cycles +P(n): 0.75 flops/cycle + diff --git a/measurements/variant_inlining_7_symmetry_2_ilp_2/arm/3.log b/measurements/variant_inlining_7_symmetry_2_ilp_2/arm/3.log new file mode 100644 index 0000000..37724e7 --- /dev/null +++ b/measurements/variant_inlining_7_symmetry_2_ilp_2/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main-symmetry.cpp ./variants/variant_inlining_7_symmetry_2_ilp_2.cpp -o main_inlining_7_symmetry_2_ilp_2 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 95500.00 flops +Q(n) per iteration: 6512 bytes +I(n): 14.67 flops/byte +W(n): 19100000.00 flops +T(n): 16525790.75 cycles +P(n): 1.16 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 371000.00 flops +Q(n) per iteration: 12912 bytes +I(n): 28.73 flops/byte +W(n): 74200000.00 flops +T(n): 65160000.00 cycles +P(n): 1.14 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 1462000.00 flops +Q(n) per iteration: 25712 bytes +I(n): 56.86 flops/byte +W(n): 292400000.00 flops +T(n): 263049074.00 cycles +P(n): 1.11 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 9055000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 141.24 flops/byte +W(n): 1811000000.00 flops +T(n): 1763111279.00 cycles +P(n): 1.03 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 36110000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 281.86 flops/byte +W(n): 7222000000.00 flops +T(n): 7502606429.00 cycles +P(n): 0.96 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 144220000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 563.11 flops/byte +W(n): 28844000000.00 flops +T(n): 35210679018.00 cycles +P(n): 0.82 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 900550000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 1406.86 flops/byte +W(n): 180110000000.00 flops +T(n): 242369036471.00 cycles +P(n): 0.74 flops/cycle + diff --git a/measurements/variant_inlining_8/arm/1.log b/measurements/variant_inlining_8/arm/1.log index cf6f049..e84af42 100644 --- a/measurements/variant_inlining_8/arm/1.log +++ b/measurements/variant_inlining_8/arm/1.log @@ -5,7 +5,7 @@ W(n) per iteration: 183700.00 flops Q(n) per iteration: 6512 bytes I(n): 28.21 flops/byte W(n): 36740000.00 flops -T(n): 15054758.12 cycles +T(n): 15034834.12 cycles P(n): 2.44 flops/cycle === N_P=100 N_ITER=200 === @@ -13,46 +13,46 @@ W(n) per iteration: 727400.00 flops Q(n) per iteration: 12912 bytes I(n): 56.34 flops/byte W(n): 145480000.00 flops -T(n): 59392245.00 cycles -P(n): 2.45 flops/cycle +T(n): 58645676.50 cycles +P(n): 2.48 flops/cycle === N_P=200 N_ITER=200 === W(n) per iteration: 2894800.00 flops Q(n) per iteration: 25712 bytes I(n): 112.59 flops/byte W(n): 578960000.00 flops -T(n): 238841666.00 cycles -P(n): 2.42 flops/cycle +T(n): 235921558.00 cycles +P(n): 2.45 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 18037000.00 flops Q(n) per iteration: 64112 bytes I(n): 281.34 flops/byte W(n): 3607400000.00 flops -T(n): 1555607186.00 cycles -P(n): 2.32 flops/cycle +T(n): 1541488178.00 cycles +P(n): 2.34 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 72074000.00 flops Q(n) per iteration: 128112 bytes I(n): 562.59 flops/byte W(n): 14414800000.00 flops -T(n): 6418988746.00 cycles -P(n): 2.25 flops/cycle +T(n): 6547730136.00 cycles +P(n): 2.20 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 288148000.00 flops Q(n) per iteration: 256112 bytes I(n): 1125.09 flops/byte W(n): 57629600000.00 flops -T(n): 25996385057.00 cycles -P(n): 2.22 flops/cycle +T(n): 26624031387.00 cycles +P(n): 2.16 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 1800370000.00 flops Q(n) per iteration: 640112 bytes I(n): 2812.59 flops/byte W(n): 360074000000.00 flops -T(n): 183377209437.00 cycles -P(n): 1.96 flops/cycle +T(n): 187629482331.00 cycles +P(n): 1.92 flops/cycle diff --git a/measurements/variant_inlining_8/arm/2.log b/measurements/variant_inlining_8/arm/2.log new file mode 100644 index 0000000..45a4428 --- /dev/null +++ b/measurements/variant_inlining_8/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_inlining_8.cpp -o main_inlining_8 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 14962674.50 cycles +P(n): 2.46 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 57946813.00 cycles +P(n): 2.51 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 232750706.00 cycles +P(n): 2.49 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 1534213485.00 cycles +P(n): 2.35 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 6376421798.00 cycles +P(n): 2.26 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 26048973254.00 cycles +P(n): 2.21 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 183555765943.00 cycles +P(n): 1.96 flops/cycle + diff --git a/measurements/variant_inlining_8/arm/3.log b/measurements/variant_inlining_8/arm/3.log new file mode 100644 index 0000000..7f1cfc1 --- /dev/null +++ b/measurements/variant_inlining_8/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_inlining_8.cpp -o main_inlining_8 -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 14940687.25 cycles +P(n): 2.46 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 57942023.00 cycles +P(n): 2.51 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 232072320.00 cycles +P(n): 2.49 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 1533490796.00 cycles +P(n): 2.35 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 6371340222.00 cycles +P(n): 2.26 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 25947418004.00 cycles +P(n): 2.22 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 183383408820.00 cycles +P(n): 1.96 flops/cycle + diff --git a/measurements/variant_lut_exp/arm/1.log b/measurements/variant_lut_exp/arm/1.log index a49fbe1..7a1ec1f 100644 --- a/measurements/variant_lut_exp/arm/1.log +++ b/measurements/variant_lut_exp/arm/1.log @@ -5,15 +5,15 @@ W(n) per iteration: 183700.00 flops Q(n) per iteration: 6512 bytes I(n): 28.21 flops/byte W(n): 36740000.00 flops -T(n): 13721813.88 cycles -P(n): 2.68 flops/cycle +T(n): 13736816.75 cycles +P(n): 2.67 flops/cycle === N_P=100 N_ITER=200 === W(n) per iteration: 727400.00 flops Q(n) per iteration: 12912 bytes I(n): 56.34 flops/byte W(n): 145480000.00 flops -T(n): 55725180.50 cycles +T(n): 55706968.00 cycles P(n): 2.61 flops/cycle === N_P=200 N_ITER=200 === @@ -21,38 +21,38 @@ W(n) per iteration: 2894800.00 flops Q(n) per iteration: 25712 bytes I(n): 112.59 flops/byte W(n): 578960000.00 flops -T(n): 213951168.00 cycles -P(n): 2.71 flops/cycle +T(n): 214494531.00 cycles +P(n): 2.70 flops/cycle === N_P=500 N_ITER=200 === W(n) per iteration: 18037000.00 flops Q(n) per iteration: 64112 bytes I(n): 281.34 flops/byte W(n): 3607400000.00 flops -T(n): 1353984010.00 cycles -P(n): 2.66 flops/cycle +T(n): 1409195434.00 cycles +P(n): 2.56 flops/cycle === N_P=1000 N_ITER=200 === W(n) per iteration: 72074000.00 flops Q(n) per iteration: 128112 bytes I(n): 562.59 flops/byte W(n): 14414800000.00 flops -T(n): 5608638594.00 cycles -P(n): 2.57 flops/cycle +T(n): 5706236811.00 cycles +P(n): 2.53 flops/cycle === N_P=2000 N_ITER=200 === W(n) per iteration: 288148000.00 flops Q(n) per iteration: 256112 bytes I(n): 1125.09 flops/byte W(n): 57629600000.00 flops -T(n): 22755683575.00 cycles -P(n): 2.53 flops/cycle +T(n): 23032405263.00 cycles +P(n): 2.50 flops/cycle === N_P=5000 N_ITER=200 === W(n) per iteration: 1800370000.00 flops Q(n) per iteration: 640112 bytes I(n): 2812.59 flops/byte W(n): 360074000000.00 flops -T(n): 151113844624.00 cycles -P(n): 2.38 flops/cycle +T(n): 150629002013.00 cycles +P(n): 2.39 flops/cycle diff --git a/measurements/variant_lut_exp/arm/2.log b/measurements/variant_lut_exp/arm/2.log new file mode 100644 index 0000000..75971b2 --- /dev/null +++ b/measurements/variant_lut_exp/arm/2.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_lut_exp.cpp -o main_lut_exp -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 13684221.25 cycles +P(n): 2.68 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 55531394.50 cycles +P(n): 2.62 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 214405137.00 cycles +P(n): 2.70 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 1355946462.00 cycles +P(n): 2.66 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 5597235102.00 cycles +P(n): 2.58 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 22733137310.00 cycles +P(n): 2.54 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 151140610676.00 cycles +P(n): 2.38 flops/cycle + diff --git a/measurements/variant_lut_exp/arm/3.log b/measurements/variant_lut_exp/arm/3.log new file mode 100644 index 0000000..91c35eb --- /dev/null +++ b/measurements/variant_lut_exp/arm/3.log @@ -0,0 +1,58 @@ +rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o +sudo /usr/bin/g++ main.cpp ./variants/variant_lut_exp.cpp -o main_lut_exp -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1 +=== N_P=50 N_ITER=200 === +W(n) per iteration: 183700.00 flops +Q(n) per iteration: 6512 bytes +I(n): 28.21 flops/byte +W(n): 36740000.00 flops +T(n): 13686852.00 cycles +P(n): 2.68 flops/cycle + +=== N_P=100 N_ITER=200 === +W(n) per iteration: 727400.00 flops +Q(n) per iteration: 12912 bytes +I(n): 56.34 flops/byte +W(n): 145480000.00 flops +T(n): 55639602.50 cycles +P(n): 2.61 flops/cycle + +=== N_P=200 N_ITER=200 === +W(n) per iteration: 2894800.00 flops +Q(n) per iteration: 25712 bytes +I(n): 112.59 flops/byte +W(n): 578960000.00 flops +T(n): 214278319.00 cycles +P(n): 2.70 flops/cycle + +=== N_P=500 N_ITER=200 === +W(n) per iteration: 18037000.00 flops +Q(n) per iteration: 64112 bytes +I(n): 281.34 flops/byte +W(n): 3607400000.00 flops +T(n): 1354936860.00 cycles +P(n): 2.66 flops/cycle + +=== N_P=1000 N_ITER=200 === +W(n) per iteration: 72074000.00 flops +Q(n) per iteration: 128112 bytes +I(n): 562.59 flops/byte +W(n): 14414800000.00 flops +T(n): 5604061102.00 cycles +P(n): 2.57 flops/cycle + +=== N_P=2000 N_ITER=200 === +W(n) per iteration: 288148000.00 flops +Q(n) per iteration: 256112 bytes +I(n): 1125.09 flops/byte +W(n): 57629600000.00 flops +T(n): 22727600983.00 cycles +P(n): 2.54 flops/cycle + +=== N_P=5000 N_ITER=200 === +W(n) per iteration: 1800370000.00 flops +Q(n) per iteration: 640112 bytes +I(n): 2812.59 flops/byte +W(n): 360074000000.00 flops +T(n): 146711076459.00 cycles +P(n): 2.45 flops/cycle + diff --git a/variants/variant_float_inlining_SoA_symmetry_ILP_16.cpp b/variants/variant_float_inlining_SoA_symmetry_ILP_16.cpp index 8e5cd0c..b64f653 100644 --- a/variants/variant_float_inlining_SoA_symmetry_ILP_16.cpp +++ b/variants/variant_float_inlining_SoA_symmetry_ILP_16.cpp @@ -7,6 +7,7 @@ #include #include +#include #define SEED 42 #define BOARD_WIDTH 10.0f diff --git a/variants/variant_float_inlining_SoA_symmetry_ILP_2.cpp b/variants/variant_float_inlining_SoA_symmetry_ILP_2.cpp index 89d55f4..dd6a429 100644 --- a/variants/variant_float_inlining_SoA_symmetry_ILP_2.cpp +++ b/variants/variant_float_inlining_SoA_symmetry_ILP_2.cpp @@ -7,6 +7,7 @@ #include #include +#include #define SEED 42 #define BOARD_WIDTH 10.0f diff --git a/variants/variant_float_inlining_SoA_symmetry_ILP_4.cpp b/variants/variant_float_inlining_SoA_symmetry_ILP_4.cpp index 3637448..e185b33 100644 --- a/variants/variant_float_inlining_SoA_symmetry_ILP_4.cpp +++ b/variants/variant_float_inlining_SoA_symmetry_ILP_4.cpp @@ -7,6 +7,7 @@ #include #include +#include #define SEED 42 #define BOARD_WIDTH 10.0f diff --git a/variants/variant_float_inlining_SoA_symmetry_ILP_8.cpp b/variants/variant_float_inlining_SoA_symmetry_ILP_8.cpp index 31fa35b..d8870cc 100644 --- a/variants/variant_float_inlining_SoA_symmetry_ILP_8.cpp +++ b/variants/variant_float_inlining_SoA_symmetry_ILP_8.cpp @@ -7,6 +7,7 @@ #include #include +#include #define SEED 42 #define BOARD_WIDTH 10.0f