SPEC2017 FPrate: LTO -Ofast - Different x86_64 levels
-flto -Ofast
AMD EPYC 7543P (zen3 based Milan)
Run-times
Benchmark
x86-64
x86-64-v2
x86-64-v3
x86-64-v2
x86-64-v3
503.bwaves_r
100.00%
99.21%
96.84%
100.00%
97.61%
507.cactuBSSN_r
100.00%
100.41%
91.84%
100.00%
91.46%
508.namd_r
100.00%
99.40%
85.74%
100.00%
86.26%
510.parest_r
100.00%
100.34%
91.44%
100.00%
91.13%
511.povray_r
100.00%
99.55%
98.87%
100.00%
99.32%
519.lbm_r
100.00%
100.00%
94.31%
100.00%
94.31%
521.wrf_r
100.00%
99.31%
85.81%
100.00%
86.41%
526.blender_r
100.00%
100.83%
100.00%
100.00%
99.18%
527.cam4_r
100.00%
97.86%
89.29%
100.00%
91.24%
538.imagick_r
100.00%
95.53%
85.11%
100.00%
89.09%
544.nab_r
100.00%
100.00%
94.75%
100.00%
94.75%
549.fotonik3d_r
100.00%
99.50%
89.85%
100.00%
90.30%
554.roms_r
100.00%
99.46%
79.09%
100.00%
79.51%
| Benchmark | x86-64 | x86-64-v2 | x86-64-v3 | x86-64-v2 | x86-64-v3 |
|---|---|---|---|---|---|
| 503.bwaves_r | 100.00% | 99.21% | 96.84% | 100.00% | 97.61% |
| 507.cactuBSSN_r | 100.00% | 100.41% | 91.84% | 100.00% | 91.46% |
| 508.namd_r | 100.00% | 99.40% | 85.74% | 100.00% | 86.26% |
| 510.parest_r | 100.00% | 100.34% | 91.44% | 100.00% | 91.13% |
| 511.povray_r | 100.00% | 99.55% | 98.87% | 100.00% | 99.32% |
| 519.lbm_r | 100.00% | 100.00% | 94.31% | 100.00% | 94.31% |
| 521.wrf_r | 100.00% | 99.31% | 85.81% | 100.00% | 86.41% |
| 526.blender_r | 100.00% | 100.83% | 100.00% | 100.00% | 99.18% |
| 527.cam4_r | 100.00% | 97.86% | 89.29% | 100.00% | 91.24% |
| 538.imagick_r | 100.00% | 95.53% | 85.11% | 100.00% | 89.09% |
| 544.nab_r | 100.00% | 100.00% | 94.75% | 100.00% | 94.75% |
| 549.fotonik3d_r | 100.00% | 99.50% | 89.85% | 100.00% | 90.30% |
| 554.roms_r | 100.00% | 99.46% | 79.09% | 100.00% | 79.51% |
Naturally, the smaller the run-times, the better. On the other hand, rates and especially their geometric means are better when they are greater.
See also continuous gcc tuning benchmarking at https://lnt.opensuse.org/db_default/v4/SPEC/spec_report/tuning.
Non-debug executable sizes
Click to display/hide
Benchmark
x86-64
x86-64-v2
x86-64-v3
x86-64-v2
x86-64-v3
503.bwaves_r
100.00%
99.78%
95.48%
100.00%
95.69%
507.cactuBSSN_r
100.00%
96.22%
99.78%
100.00%
103.70%
508.namd_r
100.00%
99.51%
99.57%
100.00%
100.07%
510.parest_r
100.00%
100.07%
104.25%
100.00%
104.18%
511.povray_r
100.00%
99.49%
97.29%
100.00%
97.79%
519.lbm_r
100.00%
99.89%
91.03%
100.00%
91.13%
521.wrf_r
100.00%
100.21%
105.49%
100.00%
105.27%
526.blender_r
100.00%
100.27%
101.95%
100.00%
101.68%
527.cam4_r
100.00%
100.10%
100.62%
100.00%
100.52%
538.imagick_r
100.00%
99.52%
103.50%
100.00%
104.00%
544.nab_r
100.00%
99.86%
100.54%
100.00%
100.68%
549.fotonik3d_r
100.00%
99.45%
117.35%
100.00%
118.00%
554.roms_r
100.00%
99.82%
108.89%
100.00%
109.09%
Smaller binary sizes are always better.
Click to display/hide
| Benchmark | x86-64 | x86-64-v2 | x86-64-v3 | x86-64-v2 | x86-64-v3 |
|---|---|---|---|---|---|
| 503.bwaves_r | 100.00% | 99.78% | 95.48% | 100.00% | 95.69% |
| 507.cactuBSSN_r | 100.00% | 96.22% | 99.78% | 100.00% | 103.70% |
| 508.namd_r | 100.00% | 99.51% | 99.57% | 100.00% | 100.07% |
| 510.parest_r | 100.00% | 100.07% | 104.25% | 100.00% | 104.18% |
| 511.povray_r | 100.00% | 99.49% | 97.29% | 100.00% | 97.79% |
| 519.lbm_r | 100.00% | 99.89% | 91.03% | 100.00% | 91.13% |
| 521.wrf_r | 100.00% | 100.21% | 105.49% | 100.00% | 105.27% |
| 526.blender_r | 100.00% | 100.27% | 101.95% | 100.00% | 101.68% |
| 527.cam4_r | 100.00% | 100.10% | 100.62% | 100.00% | 100.52% |
| 538.imagick_r | 100.00% | 99.52% | 103.50% | 100.00% | 104.00% |
| 544.nab_r | 100.00% | 99.86% | 100.54% | 100.00% | 100.68% |
| 549.fotonik3d_r | 100.00% | 99.45% | 117.35% | 100.00% | 118.00% |
| 554.roms_r | 100.00% | 99.82% | 108.89% | 100.00% | 109.09% |
Intel Cascade Lake Xeon
Run-times
Benchmark
x86-64
x86-64-v2
x86-64-v3
x86-64-v2
x86-64-v3
503.bwaves_r
100.00%
100.00%
98.29%
100.00%
98.29%
507.cactuBSSN_r
100.00%
99.58%
93.26%
100.00%
93.66%
508.namd_r
100.00%
100.00%
79.84%
100.00%
79.84%
510.parest_r
100.00%
100.00%
97.17%
100.00%
97.17%
511.povray_r
100.00%
98.61%
92.59%
100.00%
93.90%
519.lbm_r
100.00%
100.00%
97.80%
100.00%
97.80%
521.wrf_r
100.00%
99.56%
84.72%
100.00%
85.09%
526.blender_r
100.00%
97.59%
92.96%
100.00%
95.26%
527.cam4_r
100.00%
98.82%
88.42%
100.00%
89.47%
538.imagick_r
100.00%
93.58%
80.31%
100.00%
85.82%
544.nab_r
100.00%
100.00%
90.32%
100.00%
90.32%
549.fotonik3d_r
100.00%
100.66%
96.04%
100.00%
95.40%
554.roms_r
100.00%
100.52%
86.39%
100.00%
85.94%
| Benchmark | x86-64 | x86-64-v2 | x86-64-v3 | x86-64-v2 | x86-64-v3 |
|---|---|---|---|---|---|
| 503.bwaves_r | 100.00% | 100.00% | 98.29% | 100.00% | 98.29% |
| 507.cactuBSSN_r | 100.00% | 99.58% | 93.26% | 100.00% | 93.66% |
| 508.namd_r | 100.00% | 100.00% | 79.84% | 100.00% | 79.84% |
| 510.parest_r | 100.00% | 100.00% | 97.17% | 100.00% | 97.17% |
| 511.povray_r | 100.00% | 98.61% | 92.59% | 100.00% | 93.90% |
| 519.lbm_r | 100.00% | 100.00% | 97.80% | 100.00% | 97.80% |
| 521.wrf_r | 100.00% | 99.56% | 84.72% | 100.00% | 85.09% |
| 526.blender_r | 100.00% | 97.59% | 92.96% | 100.00% | 95.26% |
| 527.cam4_r | 100.00% | 98.82% | 88.42% | 100.00% | 89.47% |
| 538.imagick_r | 100.00% | 93.58% | 80.31% | 100.00% | 85.82% |
| 544.nab_r | 100.00% | 100.00% | 90.32% | 100.00% | 90.32% |
| 549.fotonik3d_r | 100.00% | 100.66% | 96.04% | 100.00% | 95.40% |
| 554.roms_r | 100.00% | 100.52% | 86.39% | 100.00% | 85.94% |
Naturally, the smaller the run-times, the better. On the other hand, rates and especially their geometric means are better when they are greater.
See also continuous gcc tuning benchmarking at https://lnt.opensuse.org/db_default/v4/SPEC/spec_report/tuning.
Non-debug executable sizes
Click to display/hide
Benchmark
x86-64
x86-64-v2
x86-64-v3
x86-64-v2
x86-64-v3
503.bwaves_r
100.00%
99.78%
95.48%
100.00%
95.69%
507.cactuBSSN_r
100.00%
96.22%
99.78%
100.00%
103.70%
508.namd_r
100.00%
99.51%
99.57%
100.00%
100.07%
510.parest_r
100.00%
100.07%
104.25%
100.00%
104.18%
511.povray_r
100.00%
99.49%
97.29%
100.00%
97.79%
519.lbm_r
100.00%
99.89%
91.03%
100.00%
91.13%
521.wrf_r
100.00%
100.21%
105.49%
100.00%
105.27%
526.blender_r
100.00%
100.27%
101.95%
100.00%
101.68%
527.cam4_r
100.00%
100.10%
100.62%
100.00%
100.52%
538.imagick_r
100.00%
99.52%
103.50%
100.00%
104.00%
544.nab_r
100.00%
99.86%
100.54%
100.00%
100.68%
549.fotonik3d_r
100.00%
99.45%
117.35%
100.00%
118.00%
554.roms_r
100.00%
99.82%
108.89%
100.00%
109.09%
Smaller binary sizes are always better.
Click to display/hide
| Benchmark | x86-64 | x86-64-v2 | x86-64-v3 | x86-64-v2 | x86-64-v3 |
|---|---|---|---|---|---|
| 503.bwaves_r | 100.00% | 99.78% | 95.48% | 100.00% | 95.69% |
| 507.cactuBSSN_r | 100.00% | 96.22% | 99.78% | 100.00% | 103.70% |
| 508.namd_r | 100.00% | 99.51% | 99.57% | 100.00% | 100.07% |
| 510.parest_r | 100.00% | 100.07% | 104.25% | 100.00% | 104.18% |
| 511.povray_r | 100.00% | 99.49% | 97.29% | 100.00% | 97.79% |
| 519.lbm_r | 100.00% | 99.89% | 91.03% | 100.00% | 91.13% |
| 521.wrf_r | 100.00% | 100.21% | 105.49% | 100.00% | 105.27% |
| 526.blender_r | 100.00% | 100.27% | 101.95% | 100.00% | 101.68% |
| 527.cam4_r | 100.00% | 100.10% | 100.62% | 100.00% | 100.52% |
| 538.imagick_r | 100.00% | 99.52% | 103.50% | 100.00% | 104.00% |
| 544.nab_r | 100.00% | 99.86% | 100.54% | 100.00% | 100.68% |
| 549.fotonik3d_r | 100.00% | 99.45% | 117.35% | 100.00% | 118.00% |
| 554.roms_r | 100.00% | 99.82% | 108.89% | 100.00% | 109.09% |