SPEC CPU result comparison

15th October 2026 - Normalized report

SPEC2026 INTrate: LTO x86-64-v3 -Ofast - Released GCCs vs Trunk

-Ofast -flto -march=x86-64-v3

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 GCC 16 Trunk
706.stockfish_r 100.00% 93.13%
707.ntest_r 100.00% 101.69%
708.sqlite_r 100.00% 100.65%
710.omnetpp_r 100.00% 96.45%
714.cpython_r 100.00% 102.36%
721.gcc_r 100.00% 100.00%
723.llvm_r 100.00% 100.00%
727.cppcheck_r 100.00% 97.84%
729.abc_r VE
734.vpr_r 100.00% 96.73%
735.gem5_r 100.00% 101.02%
750.sealcrypto_r 100.00% 45.73%
753.ns3_r 100.00% 100.00%
777.zstd_r 100.00% 100.00%
GCC 15 GCC 15.3.0 (revision 4db0e8df15bef836558857c291c323add11d035c)
GCC 16 GCC 16.2.0 (revision 78d4ac73dd391005b895a6148cd9831e28e1208b)
Trunk GCC 17.0.0 20261015 (experimental) (revision 84348add236100e5f69a3e9ec4f7b1038bf2b83d)
Naturally, the smaller the run-times, the better.
See also continuous gcc branch benchmarking at https://lnt.opensuse.org/db_default/v4/SPEC/spec_report/branch.

Non-debug executable sizes

Benchmark GCC 16 Trunk
706.stockfish_r 100.00% 99.75%
707.ntest_r 100.00% 99.87%
708.sqlite_r 100.00% 99.99%
710.omnetpp_r 100.00% 97.01%
714.cpython_r 100.00% 99.47%
721.gcc_r 100.00% 99.43%
723.llvm_r 100.00% 99.69%
727.cppcheck_r 100.00% 91.69%
729.abc_r VE
734.vpr_r 100.00% 97.52%
735.gem5_r 100.00% 99.53%
750.sealcrypto_r 100.00% 103.04%
753.ns3_r 100.00% 98.63%
777.zstd_r 100.00% 99.47%
Smaller binary sizes are always better.