Perf plot
This commit is contained in:
committed by
Linus Baumberger
parent
ca969564a6
commit
e7af3103e8
@@ -1,58 +1,66 @@
|
||||
rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o
|
||||
/usr/bin/g++ main-float.cpp ./variants/variant_float_inlining_SoA_AVX2_polyexp.cpp -o main_float_inlining_SoA_AVX2_polyexp -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1
|
||||
/usr/bin/g++ main-float.cpp ./variants/variant_float_inlining_SoA_AVX2_polyexp_expfix.cpp -o main_float_inlining_SoA_AVX2_polyexp_expfix -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): 4538634.50 cycles
|
||||
P(n): 8.09 flops/cycle
|
||||
T(n): 5860437.00 cycles
|
||||
P(n): 6.27 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): 14836382.00 cycles
|
||||
P(n): 9.81 flops/cycle
|
||||
T(n): 19736818.00 cycles
|
||||
P(n): 7.37 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): 52866688.00 cycles
|
||||
P(n): 10.95 flops/cycle
|
||||
T(n): 71295200.00 cycles
|
||||
P(n): 8.12 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): 314877088.00 cycles
|
||||
P(n): 11.46 flops/cycle
|
||||
T(n): 434200736.00 cycles
|
||||
P(n): 8.31 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): 1228216448.00 cycles
|
||||
P(n): 11.74 flops/cycle
|
||||
T(n): 1690932352.00 cycles
|
||||
P(n): 8.52 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): 4888108544.00 cycles
|
||||
P(n): 11.79 flops/cycle
|
||||
T(n): 6726832128.00 cycles
|
||||
P(n): 8.57 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): 30235383808.00 cycles
|
||||
P(n): 11.91 flops/cycle
|
||||
T(n): 41595662336.00 cycles
|
||||
P(n): 8.66 flops/cycle
|
||||
|
||||
=== N_P=10000 N_ITER=200 ===
|
||||
W(n) per iteration: 2905772800.00 flops
|
||||
Q(n) per iteration: 640056 bytes
|
||||
I(n): 4539.87 flops/byte
|
||||
W(n): 581154570240.00 flops
|
||||
T(n): 166348996608.00 cycles
|
||||
P(n): 3.49 flops/cycle
|
||||
|
||||
|
||||
@@ -1,66 +1,66 @@
|
||||
rm -f main_* run_tests *.o *.s variants/*.s variants/*.o headers/*.s headers/*.o
|
||||
/usr/bin/g++ main-float.cpp ./variants/variant_float_inlining_SoA_AVX2_polyexp.cpp -o main_float_inlining_SoA_AVX2_polyexp -O3 -march=native -std=c++23 -Wall -Wextra -flto -DFLOPCOUNT=0 -DRANDOM_SEED=0 -DROOFLINE=0 -DBENCHMARK=1
|
||||
/usr/bin/g++ main-float.cpp ./variants/variant_float_inlining_SoA_AVX2_polyexp_expfix.cpp -o main_float_inlining_SoA_AVX2_polyexp_expfix -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): 4573347.00 cycles
|
||||
P(n): 8.03 flops/cycle
|
||||
T(n): 5885449.00 cycles
|
||||
P(n): 6.24 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): 14972852.00 cycles
|
||||
P(n): 9.72 flops/cycle
|
||||
T(n): 19775262.00 cycles
|
||||
P(n): 7.36 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): 53246808.00 cycles
|
||||
P(n): 10.87 flops/cycle
|
||||
T(n): 71759640.00 cycles
|
||||
P(n): 8.07 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): 317441728.00 cycles
|
||||
P(n): 11.36 flops/cycle
|
||||
T(n): 442073856.00 cycles
|
||||
P(n): 8.16 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): 1237166592.00 cycles
|
||||
P(n): 11.65 flops/cycle
|
||||
T(n): 1698162944.00 cycles
|
||||
P(n): 8.49 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): 4899325440.00 cycles
|
||||
P(n): 11.76 flops/cycle
|
||||
T(n): 6719285248.00 cycles
|
||||
P(n): 8.58 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): 30502610944.00 cycles
|
||||
P(n): 11.80 flops/cycle
|
||||
T(n): 41795964928.00 cycles
|
||||
P(n): 8.62 flops/cycle
|
||||
|
||||
=== N_P=10000 N_ITER=200 ===
|
||||
W(n) per iteration: 2905772800.00 flops
|
||||
Q(n) per iteration: 640056 bytes
|
||||
I(n): 4539.87 flops/byte
|
||||
W(n): 581154570240.00 flops
|
||||
T(n): 122120396800.00 cycles
|
||||
P(n): 4.76 flops/cycle
|
||||
T(n): 166340608000.00 cycles
|
||||
P(n): 3.49 flops/cycle
|
||||
|
||||
|
||||
@@ -0,0 +1,64 @@
|
||||
=== 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): 5839795.00 cycles
|
||||
P(n): 6.29 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): 19706848.00 cycles
|
||||
P(n): 7.38 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): 71913632.00 cycles
|
||||
P(n): 8.05 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): 431199296.00 cycles
|
||||
P(n): 8.37 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): 1710698496.00 cycles
|
||||
P(n): 8.43 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): 6773665792.00 cycles
|
||||
P(n): 8.51 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): 43346759680.00 cycles
|
||||
P(n): 8.31 flops/cycle
|
||||
|
||||
=== N_P=10000 N_ITER=200 ===
|
||||
W(n) per iteration: 2905772800.00 flops
|
||||
Q(n) per iteration: 640056 bytes
|
||||
I(n): 4539.87 flops/byte
|
||||
W(n): 581154570240.00 flops
|
||||
T(n): 169504800768.00 cycles
|
||||
P(n): 3.43 flops/cycle
|
||||
|
||||
Reference in New Issue
Block a user