SPEC CPU result comparison

15th October 2026 - Normalized report

SPEC2026 FPrate: PGO+LTO -O2 generic - LLVM and GCC

-O2 -flto and -fprofile-instr-use vs. -profile-use

Please note that the same "optimization levels" mean different things to different compilers. For example, LLVM and LLVM-based compiler tend to be much more aggressive at -O2, often with the corresponding consequences to size of the generated binaries and compilation time. GCC is sometimes more aggressive at -O3. LLVM is also often by default less strict when it comes to standard compliance by default and performs optimizations GCC would only do with a special compilation option or at -Ofast. Keep this is mind when comparing different compilers.

AMD EPYC 8635P (zen5 based Sorano)

Run-times

Benchmark LLVM GCC
709.cactus_r 100.00% 102.43%
722.palm_r CE
731.astcenc_r 100.00% 150.48%
736.ocio_r 100.00% 100.00%
737.gmsh_r 100.00% 92.06%
748.flightdm_r 100.00% 88.17%
749.fotonik3d_r CE
765.roms_r CE
766.femflow_r 100.00% 140.00%
767.nest_r 100.00% 100.00%
772.marian_r 100.00% 253.73%
782.lbm_r 100.00% 118.37%
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)
Naturally, the smaller the run-times, the better.

Non-debug executable sizes

Benchmark LLVM GCC
709.cactus_r 100.00% 94.34%
722.palm_r CE
731.astcenc_r 100.00% 107.25%
736.ocio_r 100.00% 72.28%
737.gmsh_r 100.00% 79.62%
748.flightdm_r 100.00% 82.56%
749.fotonik3d_r CE
765.roms_r CE
766.femflow_r 100.00% 62.88%
767.nest_r 100.00% 68.17%
772.marian_r 100.00% 72.99%
782.lbm_r 100.00% 76.59%
Smaller binary sizes are always better.

AMD EPYC 75F3 (zen3 based high-frequency Milan)

Run-times

Benchmark LLVM GCC
709.cactus_r 100.00% 92.38%
722.palm_r CE
731.astcenc_r 100.00% 143.94%
736.ocio_r 100.00% 101.71%
737.gmsh_r 100.00% 92.68%
748.flightdm_r 100.00% 88.97%
749.fotonik3d_r CE
765.roms_r CE
766.femflow_r 100.00% 156.25%
767.nest_r 100.00% 100.42%
772.marian_r 100.00% 283.33%
782.lbm_r 100.00% 141.49%
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)
Naturally, the smaller the run-times, the better.

Non-debug executable sizes

Benchmark LLVM GCC
709.cactus_r 100.00% 94.34%
722.palm_r CE
731.astcenc_r 100.00% 107.25%
736.ocio_r 100.00% 71.82%
737.gmsh_r 100.00% 79.63%
748.flightdm_r 100.00% 82.30%
749.fotonik3d_r CE
765.roms_r CE
766.femflow_r 100.00% 62.70%
767.nest_r 100.00% 68.15%
772.marian_r 100.00% 72.54%
782.lbm_r 100.00% 76.83%
Smaller binary sizes are always better.