A Look At The GCC 9 Performance On Intel Skylake Against GCC 8, LLVM Clang 7/8
GCC remains much faster than Clang with the code generation performance out of the C-Ray ray-tracer.
GCC also remains faster with the binaries it produces for the LAME MP3 encoding test.
While in cases like M-Queens, the GCC lead is still there but quite small.
Clang still was pulling the occasional punches against GCC.
As we've been used to seeing in our GCC vs. Clang compiler benchmarks, the performance particularly on Intel/AMD x86_64 remains very heated and really comes down to the particular code-bases for which compiler is stronger... When counting the benchmarks where all four compilers were tested (i.e. no Fortran tests, etc), there were 32 tests in total. Clang 7.0 technically had the most wins with 9 but Clang 8.0 and GCC 9.0 were tied with eight wins each and GCC 8.2 had seven wins... All of the data is available on OpenBenchmarking.org. So basically the performance is split right along the middle and really comes down to profiling a particular code-base with each compiler option to evaluate its pros and cons.
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.