SPEC CPU result comparison

15th October 2026 - Normalized report

SPEC2017 INTrate: PGO+LTO generic -O2 - Released GCCs vs Trunk

-O2 -flto -fprofile-use

AMD EPYC 8635P (zen5 based Sorano)

Run-times

Benchmark GCC 15 GCC 16 Trunk GCC 16 Trunk
500.perlbench_r 100.00% 98.52% 100.00% 100.00% 101.50%
502.gcc_r 100.00% 98.23% 98.23% 100.00% 100.00%
505.mcf_r 100.00% 103.99% 101.60% 100.00% 97.70%
520.omnetpp_r 100.00% 99.48% 99.48% 100.00% 100.00%
523.xalancbmk_r 100.00% 95.34% 96.01% 100.00% 100.70%
525.x264_r 100.00% 92.94% 100.88% 100.00% 108.54%
531.deepsjeng_r 100.00% 102.78% 103.54% 100.00% 100.74%
541.leela_r 100.00% 98.86% 97.26% 100.00% 98.38%
548.exchange2_r 100.00% 107.98% 85.19% 100.00% 78.89%
557.xz_r 100.00% 100.00% 100.88% 100.00% 100.88%
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 15 GCC 16 Trunk GCC 16 Trunk
500.perlbench_r 100.00% 106.09% 104.15% 100.00% 98.17%
502.gcc_r 100.00% 108.74% 103.61% 100.00% 95.28%
505.mcf_r 100.00% 99.98% 100.69% 100.00% 100.71%
520.omnetpp_r 100.00% 102.04% 101.19% 100.00% 99.16%
523.xalancbmk_r 100.00% 101.78% 100.83% 100.00% 99.07%
525.x264_r 100.00% 109.85% 106.37% 100.00% 96.84%
531.deepsjeng_r 100.00% 100.18% 100.20% 100.00% 100.03%
541.leela_r 100.00% 101.11% 101.10% 100.00% 99.99%
548.exchange2_r 100.00% 98.01% 91.87% 100.00% 93.73%
557.xz_r 100.00% 102.47% 104.78% 100.00% 102.26%
Smaller binary sizes are always better.

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

Run-times

Benchmark GCC 15 GCC 16 Trunk GCC 16 Trunk
500.perlbench_r 100.00% 98.44% 98.96% 100.00% 100.53%
502.gcc_r 100.00% 97.93% 97.93% 100.00% 100.00%
505.mcf_r 100.00% 99.47% 97.88% 100.00% 98.40%
520.omnetpp_r 100.00% 98.81% 99.52% 100.00% 100.72%
523.xalancbmk_r 100.00% 96.58% 98.40% 100.00% 101.89%
525.x264_r 100.00% 94.41% 99.20% 100.00% 105.07%
531.deepsjeng_r 100.00% 101.65% 102.12% 100.00% 100.46%
541.leela_r 100.00% 97.30% 97.30% 100.00% 100.00%
548.exchange2_r 100.00% 108.36% 82.61% 100.00% 76.23%
557.xz_r 100.00% 99.56% 99.56% 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 15 GCC 16 Trunk GCC 16 Trunk
500.perlbench_r 100.00% 106.09% 104.15% 100.00% 98.17%
502.gcc_r 100.00% 108.74% 103.61% 100.00% 95.28%
505.mcf_r 100.00% 99.63% 100.69% 100.00% 101.06%
520.omnetpp_r 100.00% 102.04% 101.19% 100.00% 99.17%
523.xalancbmk_r 100.00% 101.78% 100.83% 100.00% 99.07%
525.x264_r 100.00% 109.85% 106.36% 100.00% 96.83%
531.deepsjeng_r 100.00% 100.18% 100.20% 100.00% 100.03%
541.leela_r 100.00% 101.17% 101.16% 100.00% 99.99%
548.exchange2_r 100.00% 98.01% 91.87% 100.00% 93.73%
557.xz_r 100.00% 102.44% 104.76% 100.00% 102.26%
Smaller binary sizes are always better.