The Open64 5.0 compiler was released earlier this month with many changes, among the prominently noted items were greater optimizations for AMD's Bulldozer CPUs. In this article is a first-look at the Open64 5.0 compiler performance compared to its earlier release, as tested on an AMD FX-8150 eight-core "Bulldozer" processor.
On Phoronix have already been benchmarks on AMD Bulldozer hardware comparing the performance to GCC and LLVM/Clang and Open64 Bulldozer compiler tuning, among other articles. This is the first review using the brand-new Open64 5.0 compiler release rather than Open64 4.x.
Today's testing is quite simple and similar to the earlier articles, except it is with the brand new open-source compiler. The x86_64 4.2.4 and 5.0 compiler binaries were used from the Open64 web-site. AMD has yet to release its fork of the Open64 compiler that is updated to the 5.0 status.
The two Open64 releases were tested on the FX-8150 Linux setup when using the stock configuration (not overriding any CFLAGS/CXXFLAGS from the individual test profiles) and then again when passing "-march=bdver1 -mtune=bdver1" as the compiler flags to enable the Bulldozer-specific generation-one optimizations. All testing was done by the Phoronix Test Suite