SPEC CPU result comparison

15th October 2026 - Normalized report

SPEC2026 FPrate: Standard -Ofast native - LLVM, AOCC and GCC

-Ofast -march=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 LLVM AOCC GCC AOCC GCC
709.cactus_r 100.00% 45.38% 106.43% 100.00% 234.51%
722.palm_r 100.00% 161.11% 114.05% 100.00% 70.79%
731.astcenc_r 100.00% 108.02% 116.03% 100.00% 107.42%
736.ocio_r 100.00% 97.32% 93.03% 100.00% 95.59%
737.gmsh_r 100.00% 98.55% 93.58% 100.00% 94.96%
748.flightdm_r 100.00% 100.00% 91.04% 100.00% 91.04%
749.fotonik3d_r 100.00% 100.00% 101.34% 100.00% 101.34%
765.roms_r 100.00% 101.25% 92.92% 100.00% 91.77%
766.femflow_r 100.00% VE 82.14% VE
767.nest_r 100.00% 102.82% 90.00% 100.00% 87.53%
772.marian_r 100.00% 120.00% 60.00% 100.00% 50.00%
782.lbm_r 100.00% 59.71% 61.18% 100.00% 102.46%
LLVM clang version 23.1.3 (/labs/mjambor/git/llvm 0d261d1ca552c95a8f007e061c787ac7132fbcbc), flang version 23.1.3 (/labs/mjambor/git/llvm 0d261d1ca552c95a8f007e061c787ac7132fbcbc)
AOCC AMD clang version 17.0.6 (CLANG: AOCC_5.2.0-Build#2035 2026_04_10)
GCC GCC 17.0.0 20261015 (experimental) (revision 84348add236100e5f69a3e9ec4f7b1038bf2b83d)
Naturally, the smaller the run-times, the better.

Non-debug executable sizes

Benchmark LLVM AOCC GCC AOCC GCC
709.cactus_r 100.00% 163.64% 106.65% 100.00% 65.17%
722.palm_r 100.00% 80.17% 77.29% 100.00% 96.41%
731.astcenc_r 100.00% 85.21% 137.87% 100.00% 161.80%
736.ocio_r 100.00% 92.98% 121.76% 100.00% 130.96%
737.gmsh_r 100.00% 90.15% 107.35% 100.00% 119.08%
748.flightdm_r 100.00% 98.41% 119.86% 100.00% 121.79%
749.fotonik3d_r 100.00% 37.47% 38.12% 100.00% 101.74%
765.roms_r 100.00% 72.10% 68.46% 100.00% 94.95%
766.femflow_r 100.00% VE 87.99% VE
767.nest_r 100.00% 93.89% 110.21% 100.00% 117.38%
772.marian_r 100.00% 95.55% 108.35% 100.00% 113.40%
782.lbm_r 100.00% 100.72% 52.40% 100.00% 52.02%
Smaller binary sizes are always better.