SPEC CPU result comparison

15th October 2026 - Normalized report

SPECFP 2006: PGO+LTO -Ofast native - LLVM, AOCC and GCC

-Ofast -march=native -mtune=native -flto and -fprofile-instr-use vs. -profile-use

AMD EPYC 8635P (zen5 based Sorano)

Run-times

Benchmark LLVM GCC
410.bwaves CE
416.gamess CE
433.milc 100.00% 86.85%
434.zeusmp CE
435.gromacs CE
436.cactusADM CE
437.leslie3d CE
444.namd 100.00% 114.14%
447.dealII 100.00% 97.87%
450.soplex 100.00% 78.87%
453.povray FE FE
454.calculix CE
459.GemsFDTD CE
465.tonto CE
470.lbm 100.00% 59.56%
481.wrf CE
482.sphinx3 100.00% 103.44%
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 GCC
410.bwaves CE
416.gamess CE
433.milc 100.00% 68.23%
434.zeusmp CE
435.gromacs CE
436.cactusADM CE
437.leslie3d CE
444.namd 100.00% 96.87%
447.dealII 100.00% 66.21%
450.soplex 100.00% 83.74%
453.povray FE FE
454.calculix CE
459.GemsFDTD CE
465.tonto CE
470.lbm 100.00% 57.83%
481.wrf CE
482.sphinx3 100.00% 70.08%
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% 97.71%
434.zeusmp CE
435.gromacs CE
436.cactusADM CE
437.leslie3d CE
444.namd 100.00% 109.58%
447.dealII 100.00% 98.25%
450.soplex 100.00% 102.79%
453.povray FE FE
454.calculix CE
459.GemsFDTD CE
465.tonto CE
470.lbm 100.00% 100.00%
481.wrf CE
482.sphinx3 100.00% 101.33%
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% 71.77%
434.zeusmp CE
435.gromacs CE
436.cactusADM CE
437.leslie3d CE
444.namd 100.00% 95.46%
447.dealII 100.00% 67.66%
450.soplex 100.00% 84.08%
453.povray FE FE
454.calculix CE
459.GemsFDTD CE
465.tonto CE
470.lbm 100.00% 91.97%
481.wrf CE
482.sphinx3 100.00% 70.94%
Smaller binary sizes are always better.