SPEC CPU result comparison

15th October 2026 - Normalized report

SPEC2026 INTrate: Trunk standard -Ofast - generic vs native

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

Please note that even though the header of this comparison says -Ofast, benchmarks 710.omnetpp_r, 734.vpr_r, 735.gem5_r and 753.ns3_r were additionally compiled with -fno-finite-math-only.

AMD EPYC 8635P (zen5 based Sorano)

Run-times

Benchmark Generic Native
706.stockfish_r 100.00% 64.42%
707.ntest_r 100.00% 89.08%
708.sqlite_r 100.00% 100.00%
710.omnetpp_r 100.00% 100.93%
714.cpython_r 100.00% 102.45%
721.gcc_r 100.00% 100.84%
723.llvm_r 100.00% 101.29%
727.cppcheck_r 100.00% 100.00%
729.abc_r 100.00% 99.10%
734.vpr_r 100.00% 100.00%
735.gem5_r 100.00% 101.17%
750.sealcrypto_r 100.00% 73.93%
753.ns3_r 100.00% 100.00%
777.zstd_r 100.00% 101.97%
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
706.stockfish_r 100.00% 99.60%
707.ntest_r 100.00% 100.75%
708.sqlite_r 100.00% 106.51%
710.omnetpp_r 100.00% 102.69%
714.cpython_r 100.00% 104.11%
721.gcc_r 100.00% 103.21%
723.llvm_r 100.00% 109.50%
727.cppcheck_r 100.00% 102.40%
729.abc_r 100.00% 112.73%
734.vpr_r 100.00% 104.63%
735.gem5_r 100.00% 104.97%
750.sealcrypto_r 100.00% 107.11%
753.ns3_r 100.00% 102.99%
777.zstd_r 100.00% 120.03%
Smaller binary sizes are always better.