SPEC CPU result comparison

15th October 2026 - Normalized report

SPEC2026 FPrate: Trunk standard -Ofast - generic vs native

-Ofast vs -Ofast -march=native -mtune=native

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 Generic Native
709.cactus_r 100.00% 82.81%
722.palm_r 100.00% 84.10%
731.astcenc_r 100.00% 72.90%
736.ocio_r 100.00% 75.27%
737.gmsh_r 100.00% 100.89%
748.flightdm_r 100.00% 100.00%
749.fotonik3d_r 100.00% 89.79%
765.roms_r 100.00% 71.94%
766.femflow_r 100.00% 39.66%
767.nest_r 100.00% 96.16%
772.marian_r 100.00% 36.00%
782.lbm_r 100.00% 98.11%
The compiler used was GCC 17.0.0 20261015 (experimental) (revision 84348add236100e5f69a3e9ec4f7b1038bf2b83d).
Naturally, the smaller the run-times, the better.
See also continuous gcc tuning benchmarking at https://lnt.opensuse.org/db_default/v4/SPEC/spec_report/tuning.

Non-debug executable sizes

Benchmark Generic Native
709.cactus_r 100.00% 99.56%
722.palm_r 100.00% 104.18%
731.astcenc_r 100.00% 93.91%
736.ocio_r 100.00% 104.49%
737.gmsh_r 100.00% 104.89%
748.flightdm_r 100.00% 101.23%
749.fotonik3d_r 100.00% 120.69%
765.roms_r 100.00% 104.46%
766.femflow_r 100.00% 100.37%
767.nest_r 100.00% 103.17%
772.marian_r 100.00% 103.49%
782.lbm_r 100.00% 90.23%
Smaller binary sizes are always better.