SPEC CPU result comparison

15th October 2026 - Normalized report

SPEC2026 INTrate: Trunk LTO -Ofast - x86-64-v3 vs native

-O3 -flto vs -O3 -flto and a given -march

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 x86-64-v3 Native
706.stockfish_r 100.00% 85.25%
707.ntest_r 100.00% 99.29%
708.sqlite_r 100.00% 101.52%
710.omnetpp_r 100.00% 98.87%
714.cpython_r 100.00% 98.85%
721.gcc_r 100.00% 100.64%
723.llvm_r 100.00% 99.36%
727.cppcheck_r 100.00% 101.10%
729.abc_r 100.00% 100.00%
734.vpr_r 100.00% 101.04%
735.gem5_r 100.00% 101.26%
750.sealcrypto_r 100.00% 80.83%
753.ns3_r 100.00% 100.00%
777.zstd_r 100.00% 101.42%
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 x86-64-v3 Native
706.stockfish_r 100.00% 99.09%
707.ntest_r 100.00% 101.43%
708.sqlite_r 100.00% 100.46%
710.omnetpp_r 100.00% 100.95%
714.cpython_r 100.00% 102.00%
721.gcc_r 100.00% 102.27%
723.llvm_r 100.00% 103.35%
727.cppcheck_r 100.00% 101.13%
729.abc_r 100.00% 100.85%
734.vpr_r 100.00% 101.06%
735.gem5_r 100.00% 101.27%
750.sealcrypto_r 100.00% 105.42%
753.ns3_r 100.00% 100.06%
777.zstd_r 100.00% 102.12%
Smaller binary sizes are always better.

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

Run-times

Benchmark x86-64-v3 Native
706.stockfish_r 100.00% 112.57%
707.ntest_r 100.00% 98.82%
708.sqlite_r 100.00% 98.86%
710.omnetpp_r 100.00% 100.93%
714.cpython_r 100.00% 98.33%
721.gcc_r 100.00% 100.00%
723.llvm_r 100.00% 100.00%
727.cppcheck_r 100.00% 100.00%
729.abc_r 100.00% 100.00%
734.vpr_r 100.00% 100.00%
735.gem5_r 100.00% 101.52%
750.sealcrypto_r 100.00% 98.85%
753.ns3_r 100.00% 100.77%
777.zstd_r 100.00% 100.00%
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 x86-64-v3 Native
706.stockfish_r 100.00% 99.23%
707.ntest_r 100.00% 100.11%
708.sqlite_r 100.00% 102.63%
710.omnetpp_r 100.00% 101.02%
714.cpython_r 100.00% 100.67%
721.gcc_r 100.00% 100.96%
723.llvm_r 100.00% 100.78%
727.cppcheck_r 100.00% 100.96%
729.abc_r 100.00% 100.23%
734.vpr_r 100.00% 100.61%
735.gem5_r 100.00% 100.77%
750.sealcrypto_r 100.00% 100.74%
753.ns3_r 100.00% 99.70%
777.zstd_r 100.00% 100.19%
Smaller binary sizes are always better.