LLVM Clang 6.0 vs. 5.0 Compiler Performance On Intel/AMD Linux

Several dozen benchmarks were run in looking for the C/C++ tests where there are changes to spot on any of the four Intel/AMD Linux x86_64 systems. The four systems are for a diverse look at the Clang performance but not for trying to compare the performance between the systems given the differing RAM, storage, etc.

The PolyBench-C benchmark does appear to noticeably regress for Clang 6.0 on the AMD Zen systems while the Intel performance was unchanged with this new release.

In a majority of the benchmarks, the performance of Clang 6.0 was little to no change over Clang 5.0.1, so many of those results omitted.

BLAKE2 does appear to have a minor improvement on the Intel Skylake/Coffeelake setups with Clang 6.0 while the Znver1 processors were unchanged.

In some tests, the Core i7 8700K Coffeelake system was seeing some better performance with the brand new Clang compiler update.