Skip to content
Snippets Groups Projects
Commit 3a9412e6 authored by Jan Laukemann's avatar Jan Laukemann
Browse files

initial update of benchmark runscripts

parent 5e3a72de
No related branches found
No related tags found
No related merge requests found
Showing
with 514 additions and 0 deletions
{
"compiler": {
"os": "linux",
"command": "icc",
"flags": "-Ofast -DNDEBUG -fPIC -march=native -fopenmp -std=c11",
"restrict_qualifier": "__restrict__"
},
"cache": {
"object_cache": "/home/hpc/ihpc/ihpc030h/.cache/pystencils/objectcache",
"clear_cache_on_start": false
}
}
#!/bin/sh
CC=gcc
ARMFLAGS="-mcpu=thunderx2t99 -fopenmp-simd -fno-builtin -funroll-loops"
X86FLAGS=" -fargument-noalias -funroll-loops -fno-builtin"
# Compile timing.c
$CC -O3 -c timing.c
# Compile for SKL
$CC $X86FLAGS -march=skylake-avx512 -Ofast -c copy.c
$CC $X86FLAGS -march=skylake-avx512 -Ofast -S copy.c -o copy.s.skl.s
$CC copy.o timing.o -lm -o a.copy.skl
# Compile for HSW
$CC $X86FLAGS -march=haswell -O3 -c copy.c
$CC $X86FLAGS -march=haswell -O3 -S copy.c -o copy.s.hsw.s
$CC copy.o timing.o -lm -o a.copy.hsw
# Compile for IVB
$CC $X86FLAGS -march=ivybridge -O3 -c copy.c
$CC $X86FLAGS -march=ivybridge -O3 -S copy.c -o copy.s.ivb.s
$CC copy.o timing.o -lm -o a.copy.ivb
#!/bin/bash
rm -f a.copy.*
rm -f copy.s.*
rm -f copy.o
rm -f timing.o
#include <stdio.h>
#include "timing.h"
#include <stdlib.h>
#include <math.h>
#include <unistd.h>
#include <string.h>
void dummy(double*);
void copy(long size) {
double * restrict a;
double * restrict b;
double * restrict tmp;
a = aligned_alloc(64, size*sizeof(double));
b = aligned_alloc(64, size*sizeof(double));
for(int i=0; i<size; ++i){
a[i] = 0.000000123123123123;
}
double wcs, wce, ct, runtime;
int repeat = 32;
for(runtime=0.0; runtime < 2.0; repeat *= 2) {
timing(&wcs, &ct);
for(int i=0; i<repeat; ++i) {
if(a[0] < 0) dummy(a);
//#pragma omp simd // enabled O3-like optimizations with O1
#pragma vector aligned
#pragma omp parallel for schedule(static)
for(long i=0; i < size; ++i){
b[i] = a[i];
}
tmp = a;
a = b;
b = tmp;
}
timing(&wce, &ct);
runtime = wce-wcs;
}
repeat /= 2;
printf("%12.1f | %11.8f | %9.3f | %7.1f | %7.1f | %6d | %4ld \n",
(double)size*sizeof(double)/1000.0,
runtime,
8.0*(double)size*(double)repeat*1e-6/runtime,
32.0*(double)repeat*(double)size/runtime/1024.0/1024.0,
(double)repeat*(double)size/runtime/1000.0/1000.0,
repeat,
size
);
free(a); free(b);
}
int main(int argc, char *argv[])
{
printf("COPY b[i] = a[i], 16 byte/it, 0 Flop/it\n");
printf("Size (KByte) | runtime | MFlop/s | MB/s | MLUP/s | repeat | size\n");
int size;
// 25-95
for(int i=25; i<=95; ++i) {
if( i <= 40) {
size = (int)pow(1.3, i);
} else {
size = (int)pow(1.25, i);
}
size = (size/64)*64;
copy(size);
}
return 0;
}
COPY b[i] = a[i], 16 byte/it, 0 Flop/it
Size (KByte) | runtime | MFlop/s | MB/s | MLUP/s | repeat | size
5.6 | 2.67307711 | 70697.011 | 269687.7 | 8837.1 | 33554432 | 704
7.2 | 3.37894416 | 71181.457 | 271535.7 | 8897.7 | 33554432 | 896
9.2 | 2.25411510 | 68594.023 | 261665.4 | 8574.3 | 16777216 | 1152
12.3 | 3.00619006 | 68577.976 | 261604.2 | 8572.2 | 16777216 | 1536
15.9 | 3.81385398 | 69821.229 | 266346.9 | 8727.7 | 16777216 | 1984
20.5 | 2.33323288 | 18407.795 | 70220.2 | 2301.0 | 2097152 | 2560
27.1 | 3.14568806 | 18090.896 | 69011.3 | 2261.4 | 2097152 | 3392
35.3 | 2.03855395 | 18171.750 | 69319.7 | 2271.5 | 1048576 | 4416
45.6 | 2.63785696 | 18113.761 | 69098.5 | 2264.2 | 1048576 | 5696
59.4 | 3.39013410 | 18370.077 | 70076.3 | 2296.3 | 1048576 | 7424
74.8 | 2.18155789 | 17964.949 | 68530.8 | 2245.6 | 524288 | 9344
77.3 | 2.22583485 | 18210.585 | 69467.9 | 2276.3 | 524288 | 9664
93.7 | 2.74765396 | 17878.412 | 68200.7 | 2234.8 | 524288 | 11712
100.9 | 3.01010394 | 17568.093 | 67017.0 | 2196.0 | 524288 | 12608
117.2 | 3.94125509 | 15596.991 | 59497.8 | 1949.6 | 524288 | 14656
131.1 | 2.35280681 | 14603.723 | 55708.8 | 1825.5 | 262144 | 16384
146.4 | 2.49139404 | 15407.547 | 58775.1 | 1925.9 | 262144 | 18304
170.5 | 3.67578483 | 12159.173 | 46383.6 | 1519.9 | 262144 | 21312
183.3 | 3.82256794 | 12570.070 | 47951.0 | 1571.3 | 262144 | 22912
222.2 | 2.39565802 | 12157.514 | 46377.2 | 1519.7 | 131072 | 27776
229.4 | 2.56554103 | 11718.687 | 44703.2 | 1464.8 | 131072 | 28672
286.7 | 3.30752802 | 11362.251 | 43343.5 | 1420.3 | 131072 | 35840
288.8 | 3.31905198 | 11403.678 | 43501.6 | 1425.5 | 131072 | 36096
358.4 | 2.06578803 | 11370.045 | 43373.3 | 1421.3 | 65536 | 44800
448.0 | 2.58403492 | 11362.125 | 43343.1 | 1420.3 | 65536 | 56000
560.1 | 3.23024297 | 11364.021 | 43350.3 | 1420.5 | 65536 | 70016
700.4 | 2.01982403 | 11362.986 | 43346.4 | 1420.4 | 32768 | 87552
875.5 | 2.52443910 | 11364.520 | 43352.2 | 1420.6 | 32768 | 109440
1094.7 | 3.15734601 | 11360.709 | 43337.7 | 1420.1 | 32768 | 136832
1368.1 | 3.94515300 | 11362.987 | 43346.4 | 1420.4 | 32768 | 171008
1710.1 | 2.46614099 | 11361.050 | 43339.0 | 1420.1 | 16384 | 213760
2138.1 | 3.08933306 | 11339.285 | 43255.9 | 1417.4 | 16384 | 267264
2672.6 | 3.85407186 | 11361.629 | 43341.2 | 1420.2 | 16384 | 334080
3340.8 | 2.41624308 | 11326.606 | 43207.6 | 1415.8 | 8192 | 417600
4175.9 | 3.01756287 | 11336.547 | 43245.5 | 1417.1 | 8192 | 521984
5219.8 | 3.77095199 | 11339.558 | 43257.0 | 1417.4 | 8192 | 652480
6524.9 | 2.39886308 | 11141.155 | 42500.1 | 1392.6 | 4096 | 815616
8156.2 | 3.29839110 | 10128.463 | 38637.0 | 1266.1 | 4096 | 1019520
10195.5 | 2.88422394 | 7239.484 | 27616.4 | 904.9 | 2048 | 1274432
12744.7 | 2.26356602 | 5765.494 | 21993.6 | 720.7 | 1024 | 1593088
15930.9 | 2.98102617 | 5472.351 | 20875.4 | 684.0 | 1024 | 1991360
19913.2 | 3.74652290 | 5442.682 | 20762.2 | 680.3 | 1024 | 2489152
24891.9 | 2.35192299 | 5418.823 | 20671.2 | 677.4 | 512 | 3111488
31114.8 | 2.93246603 | 5432.545 | 20723.5 | 679.1 | 512 | 3889344
38893.6 | 3.64562201 | 5462.307 | 20837.0 | 682.8 | 512 | 4861696
48617.0 | 2.28044295 | 5457.686 | 20819.4 | 682.2 | 256 | 6077120
60771.3 | 2.83332801 | 5490.879 | 20946.0 | 686.4 | 256 | 7596416
75964.4 | 3.59273601 | 5412.836 | 20648.3 | 676.6 | 256 | 9495552
94955.5 | 2.23314309 | 5442.690 | 20762.2 | 680.3 | 128 | 11869440
118694.4 | 2.80192494 | 5422.302 | 20684.4 | 677.8 | 128 | 14836800
148367.9 | 3.48113298 | 5455.433 | 20810.8 | 681.9 | 128 | 18545984
185460.2 | 2.18165803 | 5440.566 | 20754.1 | 680.1 | 64 | 23182528
231824.9 | 2.73892713 | 5417.009 | 20664.3 | 677.1 | 64 | 28978112
289781.2 | 3.41488385 | 5430.931 | 20717.4 | 678.9 | 64 | 36222656
362226.7 | 2.22380686 | 5212.347 | 19883.5 | 651.5 | 32 | 45278336
452783.6 | 2.76933098 | 5231.977 | 19958.4 | 654.0 | 32 | 56597952
565979.6 | 3.44294810 | 5260.419 | 20066.9 | 657.6 | 32 | 70747456
707474.4 | 4.31620598 | 5245.158 | 20008.7 | 655.6 | 32 | 88434304
884343.3 | 5.39312291 | 5247.235 | 20016.6 | 655.9 | 32 | 110542912
1105429.5 | 6.75115609 | 5239.657 | 19987.7 | 655.0 | 32 | 138178688
1381786.6 | 8.44252610 | 5237.434 | 19979.2 | 654.7 | 32 | 172723328
1727233.5 | 10.56710601 | 5230.521 | 19952.9 | 653.8 | 32 | 215904192
2159042.0 | 13.08003211 | 5282.047 | 20149.4 | 660.3 | 32 | 269880256
2698802.2 | 16.39192295 | 5268.550 | 20097.9 | 658.6 | 32 | 337350272
3373503.0 | 20.69668889 | 5215.911 | 19897.1 | 652.0 | 32 | 421687872
4216879.1 | 26.19448304 | 5151.471 | 19651.3 | 643.9 | 32 | 527109888
5271098.9 | 32.22887588 | 5233.666 | 19964.9 | 654.2 | 32 | 658887360
6588873.2 | 39.90079713 | 5284.204 | 20157.6 | 660.5 | 32 | 823609152
8236091.9 | 50.62682390 | 5205.836 | 19858.7 | 650.7 | 32 | 1029511488
10295114.8 | 66.18278503 | 4977.785 | 18988.7 | 622.2 | 32 | 1286889344
12868893.7 | 86.13809013 | 4780.749 | 18237.1 | 597.6 | 32 | 1608611712
COPY b[i] = a[i], 16 byte/it, 0 Flop/it
Size (KByte) | runtime | MFlop/s | MB/s | MLUP/s | repeat | size
5.6 | 2.74362493 | 34439.576 | 131376.6 | 4304.9 | 16777216 | 704
7.2 | 3.47728586 | 34584.181 | 131928.2 | 4323.0 | 16777216 | 896
9.2 | 2.29298401 | 33715.635 | 128614.9 | 4214.5 | 8388608 | 1152
12.3 | 3.26097393 | 31609.948 | 120582.4 | 3951.2 | 8388608 | 1536
15.9 | 3.88529491 | 34268.695 | 130724.7 | 4283.6 | 8388608 | 1984
20.5 | 2.78853488 | 15402.236 | 58754.9 | 1925.3 | 2097152 | 2560
27.1 | 3.78665400 | 15028.655 | 57329.8 | 1878.6 | 2097152 | 3392
35.3 | 2.45238113 | 15105.357 | 57622.4 | 1888.2 | 1048576 | 4416
45.6 | 3.15658617 | 15137.084 | 57743.4 | 1892.1 | 1048576 | 5696
59.4 | 2.01748395 | 15434.330 | 58877.3 | 1929.3 | 524288 | 7424
74.8 | 2.60967994 | 15017.771 | 57288.3 | 1877.2 | 524288 | 9344
77.3 | 2.78718591 | 14542.896 | 55476.7 | 1817.9 | 524288 | 9664
93.7 | 3.50938797 | 13997.794 | 53397.3 | 1749.7 | 524288 | 11712
100.9 | 3.76809502 | 14034.090 | 53535.8 | 1754.3 | 524288 | 12608
117.2 | 2.29294705 | 13404.522 | 51134.2 | 1675.6 | 262144 | 14656
131.1 | 2.57975507 | 13318.992 | 50807.9 | 1664.9 | 262144 | 16384
146.4 | 3.02592683 | 12685.789 | 48392.4 | 1585.7 | 262144 | 18304
170.5 | 3.71225309 | 12039.724 | 45927.9 | 1505.0 | 262144 | 21312
183.3 | 2.17296815 | 11056.293 | 42176.4 | 1382.0 | 131072 | 22912
222.2 | 3.10106802 | 9392.005 | 35827.7 | 1174.0 | 131072 | 27776
229.4 | 3.28877902 | 9141.621 | 34872.5 | 1142.7 | 131072 | 28672
286.7 | 2.08392000 | 9016.892 | 34396.7 | 1127.1 | 65536 | 35840
288.8 | 2.08808017 | 9063.205 | 34573.4 | 1132.9 | 65536 | 36096
358.4 | 2.59581494 | 9048.450 | 34517.1 | 1131.1 | 65536 | 44800
448.0 | 3.24463201 | 9048.831 | 34518.6 | 1131.1 | 65536 | 56000
560.1 | 2.02827001 | 9049.226 | 34520.1 | 1131.2 | 32768 | 70016
700.4 | 2.53634191 | 9048.950 | 34519.0 | 1131.1 | 32768 | 87552
875.5 | 3.17099905 | 9047.319 | 34512.8 | 1130.9 | 32768 | 109440
1094.7 | 3.97460413 | 9024.720 | 34426.6 | 1128.1 | 32768 | 136832
1368.1 | 2.50004411 | 8965.586 | 34201.0 | 1120.7 | 16384 | 171008
1710.1 | 3.12586594 | 8963.261 | 34192.1 | 1120.4 | 16384 | 213760
2138.1 | 3.90748596 | 8965.055 | 34199.0 | 1120.6 | 16384 | 267264
2672.6 | 2.43022895 | 9009.138 | 34367.1 | 1126.1 | 8192 | 334080
3340.8 | 3.03376508 | 9021.079 | 34412.7 | 1127.6 | 8192 | 417600
4175.9 | 3.79373312 | 9017.172 | 34397.8 | 1127.1 | 8192 | 521984
5219.8 | 2.36844087 | 9027.232 | 34436.2 | 1128.4 | 4096 | 652480
6524.9 | 2.95556712 | 9042.632 | 34494.9 | 1130.3 | 4096 | 815616
8156.2 | 3.71996999 | 8980.619 | 34258.3 | 1122.6 | 4096 | 1019520
10195.5 | 2.42400408 | 8613.968 | 32859.7 | 1076.7 | 2048 | 1274432
12744.7 | 3.70229912 | 7049.985 | 26893.6 | 881.2 | 2048 | 1593088
15930.9 | 3.05246997 | 5344.269 | 20386.8 | 668.0 | 1024 | 1991360
19913.2 | 2.13063908 | 4785.215 | 18254.1 | 598.2 | 512 | 2489152
24891.9 | 2.76694512 | 4606.038 | 17570.6 | 575.8 | 512 | 3111488
31114.8 | 3.38911700 | 4700.562 | 17931.2 | 587.6 | 512 | 3889344
38893.6 | 2.12151504 | 4693.228 | 17903.2 | 586.7 | 256 | 4861696
48617.0 | 2.71741080 | 4580.074 | 17471.6 | 572.5 | 256 | 6077120
60771.3 | 3.30090308 | 4713.092 | 17979.0 | 589.1 | 256 | 7596416
75964.4 | 2.07672310 | 4682.110 | 17860.8 | 585.3 | 128 | 9495552
94955.5 | 2.59514308 | 4683.482 | 17866.1 | 585.4 | 128 | 11869440
118694.4 | 3.24267220 | 4685.297 | 17873.0 | 585.7 | 128 | 14836800
148367.9 | 2.07651997 | 4572.816 | 17443.9 | 571.6 | 64 | 18545984
185460.2 | 2.56732512 | 4623.277 | 17636.4 | 577.9 | 64 | 23182528
231824.9 | 3.16727090 | 4684.409 | 17869.6 | 585.6 | 64 | 28978112
289781.2 | 2.08523107 | 4446.989 | 16963.9 | 555.9 | 32 | 36222656
362226.7 | 2.54679108 | 4551.317 | 17361.9 | 568.9 | 32 | 45278336
452783.6 | 3.18031812 | 4555.857 | 17379.2 | 569.5 | 32 | 56597952
565979.6 | 4.07805896 | 4441.169 | 16941.7 | 555.1 | 32 | 70747456
707474.4 | 4.95661211 | 4567.471 | 17423.5 | 570.9 | 32 | 88434304
884343.3 | 6.32752609 | 4472.362 | 17060.7 | 559.0 | 32 | 110542912
1105429.5 | 7.80907607 | 4529.825 | 17279.9 | 566.2 | 32 | 138178688
1381786.6 | 9.73703694 | 4541.132 | 17323.0 | 567.6 | 32 | 172723328
1727233.5 | 12.07400417 | 4577.725 | 17462.6 | 572.2 | 32 | 215904192
2159042.0 | 15.18957591 | 4548.471 | 17351.0 | 568.6 | 32 | 269880256
2698802.2 | 18.86034799 | 4579.007 | 17467.5 | 572.4 | 32 | 337350272
3373503.0 | 23.83696389 | 4528.769 | 17275.9 | 566.1 | 32 | 421687872
4216879.1 | 29.46909595 | 4579.039 | 17467.6 | 572.4 | 32 | 527109888
5271098.9 | 36.75355792 | 4589.356 | 17507.0 | 573.7 | 32 | 658887360
6588873.2 | 46.68279314 | 4516.524 | 17229.2 | 564.6 | 32 | 823609152
8236091.9 | 57.60187411 | 4575.458 | 17454.0 | 571.9 | 32 | 1029511488
10295114.8 | 71.86370206 | 4584.285 | 17487.7 | 573.0 | 32 | 1286889344
12868893.7 | 91.67534995 | 4491.988 | 17135.6 | 561.5 | 32 | 1608611712
COPY b[i] = a[i], 16 byte/it, 0 Flop/it
Size (KByte) | runtime | MFlop/s | MB/s | MLUP/s | repeat | size
5.6 | 3.08191299 | 122637.181 | 467823.7 | 15329.6 | 67108864 | 704
7.2 | 3.69210696 | 130287.758 | 497008.4 | 16286.0 | 67108864 | 896
9.2 | 2.55285501 | 121134.042 | 462089.7 | 15141.8 | 33554432 | 1152
12.3 | 3.29275203 | 125219.530 | 477674.6 | 15652.4 | 33554432 | 1536
15.9 | 2.10585594 | 126451.182 | 482373.0 | 15806.4 | 16777216 | 1984
20.5 | 2.39175296 | 35914.807 | 137004.1 | 4489.4 | 4194304 | 2560
27.1 | 3.45209098 | 32970.346 | 125771.9 | 4121.3 | 4194304 | 3392
35.3 | 2.22132087 | 33353.212 | 127232.4 | 4169.2 | 2097152 | 4416
45.6 | 2.82396698 | 33839.993 | 129089.3 | 4230.0 | 2097152 | 5696
59.4 | 3.56987810 | 34890.281 | 133095.9 | 4361.3 | 2097152 | 7424
74.8 | 2.39221692 | 32765.905 | 124992.0 | 4095.7 | 1048576 | 9344
77.3 | 2.36496782 | 34278.482 | 130762.0 | 4284.8 | 1048576 | 9664
93.7 | 2.84613514 | 34519.576 | 131681.7 | 4314.9 | 1048576 | 11712
100.9 | 3.11853004 | 33914.559 | 129373.8 | 4239.3 | 1048576 | 12608
117.2 | 3.62759495 | 33891.170 | 129284.6 | 4236.4 | 1048576 | 14656
131.1 | 2.02053189 | 34010.588 | 129740.1 | 4251.3 | 524288 | 16384
146.4 | 2.36161399 | 32508.505 | 124010.1 | 4063.6 | 524288 | 18304
170.5 | 2.53340197 | 35284.178 | 134598.5 | 4410.5 | 524288 | 21312
183.3 | 2.93269801 | 32768.424 | 125001.6 | 4096.1 | 524288 | 22912
222.2 | 3.28996301 | 35411.033 | 135082.4 | 4426.4 | 524288 | 27776
229.4 | 3.43761706 | 34983.270 | 133450.6 | 4372.9 | 524288 | 28672
286.7 | 2.15969610 | 34802.085 | 132759.4 | 4350.3 | 262144 | 35840
288.8 | 2.15335703 | 35153.854 | 134101.3 | 4394.2 | 262144 | 36096
358.4 | 2.75105906 | 34151.361 | 130277.1 | 4268.9 | 262144 | 44800
448.0 | 2.33018994 | 25199.772 | 96129.5 | 3150.0 | 131072 | 56000
560.1 | 3.65149593 | 20106.033 | 76698.4 | 2513.3 | 131072 | 70016
700.4 | 3.31449103 | 13849.023 | 52829.8 | 1731.1 | 65536 | 87552
875.5 | 2.37142801 | 12097.791 | 46149.4 | 1512.2 | 32768 | 109440
1094.7 | 3.06186199 | 11714.992 | 44689.1 | 1464.4 | 32768 | 136832
1368.1 | 3.83201003 | 11698.487 | 44626.2 | 1462.3 | 32768 | 171008
1710.1 | 2.39518213 | 11697.628 | 44622.9 | 1462.2 | 16384 | 213760
2138.1 | 2.99354601 | 11702.117 | 44640.0 | 1462.8 | 16384 | 267264
2672.6 | 3.74240398 | 11700.643 | 44634.4 | 1462.6 | 16384 | 334080
3340.8 | 2.33919120 | 11699.699 | 44630.8 | 1462.5 | 8192 | 417600
4175.9 | 2.92346787 | 11701.426 | 44637.4 | 1462.7 | 8192 | 521984
5219.8 | 3.65539503 | 11698.032 | 44624.5 | 1462.3 | 8192 | 652480
6524.9 | 2.28962493 | 11672.700 | 44527.8 | 1459.1 | 4096 | 815616
8156.2 | 2.87826896 | 11606.848 | 44276.6 | 1450.9 | 4096 | 1019520
10195.5 | 3.98002481 | 10492.545 | 40025.9 | 1311.6 | 4096 | 1274432
12744.7 | 2.61160898 | 9994.281 | 38125.2 | 1249.3 | 2048 | 1593088
15930.9 | 3.51269889 | 9288.141 | 35431.4 | 1161.0 | 2048 | 1991360
19913.2 | 2.35897207 | 8644.076 | 32974.5 | 1080.5 | 1024 | 2489152
24891.9 | 3.11050200 | 8194.597 | 31259.9 | 1024.3 | 1024 | 3111488
31114.8 | 2.00837493 | 7932.161 | 30258.8 | 991.5 | 512 | 3889344
38893.6 | 2.59061909 | 7686.775 | 29322.7 | 960.8 | 512 | 4861696
48617.0 | 3.32759309 | 7480.447 | 28535.6 | 935.1 | 512 | 6077120
60771.3 | 2.08625913 | 7457.108 | 28446.6 | 932.1 | 256 | 7596416
75964.4 | 2.63398385 | 7383.071 | 28164.2 | 922.9 | 256 | 9495552
94955.5 | 3.32092214 | 7319.838 | 27923.0 | 915.0 | 256 | 11869440
118694.4 | 2.07402086 | 7325.328 | 27943.9 | 915.7 | 128 | 14836800
148367.9 | 2.60452199 | 7291.583 | 27815.2 | 911.4 | 128 | 18545984
185460.2 | 3.27135301 | 7256.603 | 27681.7 | 907.1 | 128 | 23182528
231824.9 | 2.04595613 | 7251.765 | 27663.3 | 906.5 | 64 | 28978112
289781.2 | 2.56240106 | 7237.743 | 27609.8 | 904.7 | 64 | 36222656
362226.7 | 3.20592809 | 7231.138 | 27584.6 | 903.9 | 64 | 45278336
452783.6 | 2.06973481 | 7000.450 | 26704.6 | 875.1 | 32 | 56597952
565979.6 | 2.58132815 | 7016.291 | 26765.0 | 877.0 | 32 | 70747456
707474.4 | 3.23389602 | 7000.591 | 26705.1 | 875.1 | 32 | 88434304
884343.3 | 4.05660486 | 6976.027 | 26611.4 | 872.0 | 32 | 110542912
1105429.5 | 5.09171486 | 6947.314 | 26501.9 | 868.4 | 32 | 138178688
1381786.6 | 6.35625410 | 6956.483 | 26536.9 | 869.6 | 32 | 172723328
1727233.5 | 7.94153500 | 6959.797 | 26549.5 | 870.0 | 32 | 215904192
2159042.0 | 9.89042187 | 6985.480 | 26647.5 | 873.2 | 32 | 269880256
2698802.2 | 12.48265886 | 6918.532 | 26392.1 | 864.8 | 32 | 337350272
3373503.0 | 15.47786808 | 6974.610 | 26606.0 | 871.8 | 32 | 421687872
4216879.1 | 19.28635192 | 6996.664 | 26690.2 | 874.6 | 32 | 527109888
5271098.9 | 24.13399792 | 6989.110 | 26661.3 | 873.6 | 32 | 658887360
6588873.2 | 30.50067091 | 6912.764 | 26370.1 | 864.1 | 32 | 823609152
8236091.9 | 38.15790606 | 6906.955 | 26347.9 | 863.4 | 32 | 1029511488
10295114.8 | 47.83115005 | 6887.639 | 26274.3 | 861.0 | 32 | 1286889344
12868893.7 | 59.28840494 | 6945.786 | 26496.1 | 868.2 | 32 | 1608611712
#!/bin/bash
ARCH=`echo $1 | awk '{print tolower($0)}'`
ARCH_UP=`echo $1 | awk '{print toupper($0)}'`
CSV_FILE="../results_copy.csv"
if [ $# -eq 0 ]; then
echo "Specify target architecture [IVB, HSW, SKL]"
exit -1
fi
if [ $1 = "--help" ]; then
echo -e "Usage: ./run_copy.sh [ARCH]\n\n ARCH Target architecture. One of [IVB, HSW, SKL].\n"
exit 0
fi
# Build benchmark
./build.sh
# Run benchmark
./a.copy.$ARCH | sort -nk1 | tee -a out.copy.$ARCH
# Store result in csv for plot
if [ ! -f $CSV_FILE ]; then
echo "size(KB),bandwidth(MB/s),arch" > $CSV_FILE
fi
awk -v arch="$ARCH_UP" 'NR>2 {print $1, ",", $5, ",", arch}' out.copy.$ARCH >> $CSV_FILE
#include "timing.h"
void timing(double* wcTime, double* cpuTime)
{
struct timeval tp;
struct rusage ruse;
gettimeofday(&tp, NULL);
*wcTime=(double) (tp.tv_sec + tp.tv_usec/1000000.0);
getrusage(RUSAGE_SELF, &ruse);
*cpuTime=(double)(ruse.ru_utime.tv_sec+ruse.ru_utime.tv_usec / 1000000.0);
}
void timing_(double* wcTime, double* cpuTime) {
timing(wcTime, cpuTime);
}
void dummy(double* a) {}
#include <sys/time.h>
#include <sys/types.h>
#include <stdlib.h>
#include <stdio.h>
#include <sys/resource.h>
void timing(double* wcTime, double* cpuTime);
File added
File added
File added
File added
File added
File added
File added
#!/bin/bash
rm -f density-ux.dat
rm -f ux-progress.dat
#!/bin/bash
# Check parameter
while test $# -gt 0; do
case "$1" in
-h|--help)
echo "Usage: ./create_plotdata_lbmBench_node_scale.sh ARCH"
echo -e "\n ARCH Target architecture. One of [IVB, HSW, SKL]\n"
exit 0
;;
*)
ARCH=`echo $1 | awk '{print tolower($0)}'`
shift
;;
esac
done
if [ -z $ARCH ]; then
echo "Specify the target architecture [SKL, HSW, IVB]."
exit -1
fi
case $ARCH in
"skl")
FLAGS="avx512"
# THREADS="1-40"
;;
"hsw")
FLAGS="avx2"
# THREADS="1-28"
;;
"ivb")
FLAGS="avx"
# THREADS="1-20"
;;
*)
echo "Target architecture '$ARCH' not supported."
exit -1
;;
esac
##########################
CSV_FILE="../results_lbmBench_node.csv"
CSV_FILE_ALL="../results_node.csv"
BENCH_OUT="out.lbmBench.$ARCH"
THREADS="1-40"
# Prepare CSV
if [ ! -f $CSV_FILE ]; then
echo "MFLUPS,threads,dim_x,kernel" > $CSV_FILE
fi
if [ ! -f $CSV_FILE_ALL ]; then
echo "MFLUPS,threads,dim_x,kernel" > $CSV_FILE_ALL
fi
for flag in $FLAGS
do
# Run benchmarks
./run_lbmBench-kernels.sh $ARCH $flag -t $THREADS -d 300 -f
done
# Format output
cat $BENCH_OUT | awk -v arch="$ARCH" '{OFS=","; print $3, $2, $4, toupper(arch)"_"$5}' >> $CSV_FILE
cat $BENCH_OUT | awk -v arch="$ARCH" '{OFS=","; print $3, $2, $4, toupper(arch)"_"$5}' >> $CSV_FILE_ALL
#!/bin/bash
# Check parameter
while test $# -gt 0; do
case "$1" in
-h|--help)
echo "Usage: ./create_plotdata_lbmBench_single_core.sh ARCH"
echo -e "\n ARCH Target architecture. One of [IVB, HSW, SKL]\n"
exit 0
;;
*)
ARCH=`echo $1 | awk '{print tolower($0)}'`
shift
;;
esac
done
if [ -z $ARCH ]; then
echo "Specify the target architecture [SKL, HSW, IVB]."
exit -1
fi
case $ARCH in
"skl")
FLAGS="avx512 avx2 avx"
;;
"hsw")
FLAGS="avx2 avx"
;;
"ivb")
FLAGS="avx"
;;
*)
echo "Target architecture '$ARCH' not supported."
exit -1
;;
esac
##########################
CSV_FILE="../results_lbmBench_$ARCH.csv"
BENCH_OUT="out.lbmBench.$ARCH"
# Prepare CSV
echo "flags,MFLUPS,threads,kernel" > $CSV_FILE
for flag in $FLAGS
do
# Run benchmarks
./run_lbmBench-kernels.sh $ARCH $flag -t 1 -d 300
done
# Format output
cat $BENCH_OUT | awk '{OFS=","; print $1, $3, $2, $5}' >> $CSV_FILE
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment