Intel Xeon Skylake Compilers: Clang Showing Strong Performance Against GCC

A few days ago on the new Intel Xeon E3 1245 v5 "Skylake" system I ran a variety of GCC and LLVM Clang compiler benchmarks to show how the performance of the resulting binaries differ between these competing open-source compilers.

From the Xeon E3 Skylake system running Ubuntu 15.10 Linux, the following compilers were tested:

- GCC 4.9.3
- GCC 5.2.1
- GCC 6.0 from its 6 December snapshot.
- LLVM Clang 3.8 from its SVN code-base as of last week. OpenMP support is enabled.

Ubuntu 15.10 x86_64 was running on this system while having manually switched to the Linux 4.4 Git kernel for all of this testing in order to provide the best Intel Skylake support.

All of the same CFLAGS/CXXFLAGS were maintained during the benchmarking process. All of the compilers tested were in their released/non-debug modes. All of these compiler benchmarks were done in a fully-automated and reproducible manner using the open-source Phoronix Test Suite benchmarking software.