SPEC2026 FPrate: LTO -Ofast native - LLVM and GCC
-Ofast -march=native -flto
Please note that event though the header of this comparison
says -Ofast, benchmark 767.nest_r was additionally compiled
with -fno-finite-math-only. Furthermore, benchmarks
737.gmsh_r and 748.flightdm_r were compiled
with -fno-fast-math, so a big chunk of -Ofast was
in fact disabled. (This might put GCC at a disadvantage compared to
LLVM-based compilers because the latter defaults
to -fno-math-errno and -fno-trapping-math, in
future measurements, we will try to enable these flags.)
AMD EPYC 8635P (zen5 based Sorano)
Run-times
Benchmark
LLVM
GCC
709.cactus_r
100.00%
115.26%
722.palm_r
100.00%
120.34%
731.astcenc_r
100.00%
117.18%
736.ocio_r
100.00%
94.04%
737.gmsh_r
100.00%
93.28%
748.flightdm_r
100.00%
92.78%
749.fotonik3d_r
100.00%
102.68%
765.roms_r
100.00%
93.25%
766.femflow_r
100.00%
95.92%
767.nest_r
100.00%
98.08%
772.marian_r
100.00%
64.29%
782.lbm_r
100.00%
60.23%
| Benchmark | LLVM | GCC |
|---|---|---|
| 709.cactus_r | 100.00% | 115.26% |
| 722.palm_r | 100.00% | 120.34% |
| 731.astcenc_r | 100.00% | 117.18% |
| 736.ocio_r | 100.00% | 94.04% |
| 737.gmsh_r | 100.00% | 93.28% |
| 748.flightdm_r | 100.00% | 92.78% |
| 749.fotonik3d_r | 100.00% | 102.68% |
| 765.roms_r | 100.00% | 93.25% |
| 766.femflow_r | 100.00% | 95.92% |
| 767.nest_r | 100.00% | 98.08% |
| 772.marian_r | 100.00% | 64.29% |
| 782.lbm_r | 100.00% | 60.23% |
| LLVM | clang version 23.1.3 (/labs/mjambor/git/llvm 0d261d1ca552c95a8f007e061c787ac7132fbcbc), flang version 23.1.3 (/labs/mjambor/git/llvm 0d261d1ca552c95a8f007e061c787ac7132fbcbc) |
| GCC | GCC 17.0.0 20261015 (experimental) (revision 84348add236100e5f69a3e9ec4f7b1038bf2b83d) |
Non-debug executable sizes
Benchmark
LLVM
GCC
709.cactus_r
100.00%
95.13%
722.palm_r
100.00%
82.60%
731.astcenc_r
100.00%
138.64%
736.ocio_r
100.00%
90.22%
737.gmsh_r
100.00%
96.65%
748.flightdm_r
100.00%
91.45%
749.fotonik3d_r
100.00%
35.38%
765.roms_r
100.00%
70.58%
766.femflow_r
100.00%
67.62%
767.nest_r
100.00%
84.99%
772.marian_r
100.00%
93.60%
782.lbm_r
100.00%
45.12%
Smaller binary sizes are always better.
| Benchmark | LLVM | GCC |
|---|---|---|
| 709.cactus_r | 100.00% | 95.13% |
| 722.palm_r | 100.00% | 82.60% |
| 731.astcenc_r | 100.00% | 138.64% |
| 736.ocio_r | 100.00% | 90.22% |
| 737.gmsh_r | 100.00% | 96.65% |
| 748.flightdm_r | 100.00% | 91.45% |
| 749.fotonik3d_r | 100.00% | 35.38% |
| 765.roms_r | 100.00% | 70.58% |
| 766.femflow_r | 100.00% | 67.62% |
| 767.nest_r | 100.00% | 84.99% |
| 772.marian_r | 100.00% | 93.60% |
| 782.lbm_r | 100.00% | 45.12% |
AMD EPYC 75F3 (zen3 based high-frequency Milan)
Run-times
Benchmark
LLVM
GCC
709.cactus_r
100.00%
100.00%
722.palm_r
100.00%
127.24%
731.astcenc_r
100.00%
103.99%
736.ocio_r
100.00%
92.53%
737.gmsh_r
100.00%
95.14%
748.flightdm_r
100.00%
94.69%
749.fotonik3d_r
100.00%
104.28%
765.roms_r
100.00%
92.62%
766.femflow_r
100.00%
117.19%
767.nest_r
100.00%
101.88%
772.marian_r
100.00%
116.00%
782.lbm_r
100.00%
114.56%
| Benchmark | LLVM | GCC |
|---|---|---|
| 709.cactus_r | 100.00% | 100.00% |
| 722.palm_r | 100.00% | 127.24% |
| 731.astcenc_r | 100.00% | 103.99% |
| 736.ocio_r | 100.00% | 92.53% |
| 737.gmsh_r | 100.00% | 95.14% |
| 748.flightdm_r | 100.00% | 94.69% |
| 749.fotonik3d_r | 100.00% | 104.28% |
| 765.roms_r | 100.00% | 92.62% |
| 766.femflow_r | 100.00% | 117.19% |
| 767.nest_r | 100.00% | 101.88% |
| 772.marian_r | 100.00% | 116.00% |
| 782.lbm_r | 100.00% | 114.56% |
| LLVM | clang version 23.1.3 (https://github.com/llvm/llvm-project.git 0d261d1ca552c95a8f007e061c787ac7132fbcbc), flang version 23.1.3 (https://github.com/llvm/llvm-project.git 0d261d1ca552c95a8f007e061c787ac7132fbcbc) |
| GCC | GCC 17.0.0 20261015 (experimental) (revision 84348add236100e5f69a3e9ec4f7b1038bf2b83d) |
Non-debug executable sizes
Benchmark
LLVM
GCC
709.cactus_r
100.00%
96.35%
722.palm_r
100.00%
83.41%
731.astcenc_r
100.00%
160.42%
736.ocio_r
100.00%
91.86%
737.gmsh_r
100.00%
97.57%
748.flightdm_r
100.00%
91.77%
749.fotonik3d_r
100.00%
36.03%
765.roms_r
100.00%
71.87%
766.femflow_r
100.00%
67.11%
767.nest_r
100.00%
87.55%
772.marian_r
100.00%
92.87%
782.lbm_r
100.00%
70.28%
Smaller binary sizes are always better.
| Benchmark | LLVM | GCC |
|---|---|---|
| 709.cactus_r | 100.00% | 96.35% |
| 722.palm_r | 100.00% | 83.41% |
| 731.astcenc_r | 100.00% | 160.42% |
| 736.ocio_r | 100.00% | 91.86% |
| 737.gmsh_r | 100.00% | 97.57% |
| 748.flightdm_r | 100.00% | 91.77% |
| 749.fotonik3d_r | 100.00% | 36.03% |
| 765.roms_r | 100.00% | 71.87% |
| 766.femflow_r | 100.00% | 67.11% |
| 767.nest_r | 100.00% | 87.55% |
| 772.marian_r | 100.00% | 92.87% |
| 782.lbm_r | 100.00% | 70.28% |