SPEC CPU result comparison

15th October 2026 - Normalized report

SPECFP 2006: 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
410.bwaves CE
416.gamess CE
433.milc 100.00% 98.02%
434.zeusmp CE
435.gromacs CE
436.cactusADM CE
437.leslie3d CE
444.namd 100.00% 102.09%
447.dealII 100.00% 98.96%
450.soplex 100.00% 101.07%
453.povray 100.00% 95.13%
454.calculix CE
459.GemsFDTD CE
465.tonto CE
470.lbm 100.00% 46.89%
481.wrf CE
482.sphinx3 100.00% 95.38%
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
410.bwaves CE
416.gamess CE
433.milc 100.00% 86.16%
434.zeusmp CE
435.gromacs CE
436.cactusADM CE
437.leslie3d CE
444.namd 100.00% 111.08%
447.dealII 100.00% 76.42%
450.soplex 100.00% 91.55%
453.povray 100.00% 88.10%
454.calculix CE
459.GemsFDTD CE
465.tonto CE
470.lbm 100.00% 85.08%
481.wrf CE
482.sphinx3 100.00% 80.96%
Smaller binary sizes are always better.

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

Run-times

Benchmark LLVM GCC
410.bwaves CE
416.gamess CE
433.milc 100.00% 75.20%
434.zeusmp CE
435.gromacs CE
436.cactusADM CE
437.leslie3d CE
444.namd 100.00% 99.39%
447.dealII 100.00% 103.35%
450.soplex 100.00% 100.00%
453.povray 100.00% 95.05%
454.calculix CE
459.GemsFDTD CE
465.tonto CE
470.lbm 100.00% 45.35%
481.wrf CE
482.sphinx3 100.00% 94.55%
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
410.bwaves CE
416.gamess CE
433.milc 100.00% 86.15%
434.zeusmp CE
435.gromacs CE
436.cactusADM CE
437.leslie3d CE
444.namd 100.00% 111.09%
447.dealII 100.00% 76.68%
450.soplex 100.00% 91.56%
453.povray 100.00% 88.10%
454.calculix CE
459.GemsFDTD CE
465.tonto CE
470.lbm 100.00% 85.39%
481.wrf CE
482.sphinx3 100.00% 80.99%
Smaller binary sizes are always better.