AMD AOCC 4.0 vs. GCC vs. LLVM Clang Compiler Benchmarks On Zen 4
I carried out a wide variety of C/C++ benchmarks built under these five different compilers. See all of my benchmark results in full via this result page.
Out of 190 tests carried out, the AMD Optimizing C/C++ Compiler was the front-runner around 55% of the time. The rest of the time it was an almost even split between the other LLVM/Clang and GCC compilers/versions tested.
The AOCC 4.0 built binaries were the slowest just 9.5% of the time while GCC 12.2 had the most last place finishes at 40%.
If taking the geometric mean for the subset of benchmarks that all ran successfully on all tested compilers, AOCC 4.0 was the exception for its performance being measurably different than the rest. AOCC 4.0 overall was yielding around 4% better performance than the current GCC 12 stable release. Of course, it will be interesting to see how these numbers compare once both GCC and LLVM/Clang upstream have had optimized Zen 4 support and tuning take place. In any event, AOCC 4.0 shows the potential and benefit to compiler optimizations and for software vendors and developers it certainly can pay off to evaluate the performance of alternative compilers. This round of benchmarking was with an AMD Ryzen 9 7950X while as system time allows will also carry out even more compiler benchmarks on a high core count AMD EPYC "Genoa" server.
If you enjoyed this article consider joining Phoronix Premium to view this site ad-free, multi-page articles on a single page, and other benefits. PayPal or Stripe tips are also graciously accepted. Thanks for your support.