AMD AOCC 1.3 Compiler Benchmarks vs. GCC 8.2 vs. LLVM Clang 7.0
The High Performance Conjugate Gradient (HPCG) was first up seeing a rise in performance compared to LLVM Clang 7.0, which also came in at the same speed as GCC 8.2.
PolyBench-C was also yielding the fastest performance when built by AOCC 1.3, which was slightly faster than GCC 8.2.0 and noticeably better than vanilla LLVM Clang 7.
The FFTW results meanwhile were mixed with a tight race between GCC 8.2 and AOCC 1.3 while Clang 7.0 stayed in last place.
But in the HMMer benchmark, LLVM Clang 7.0 upstream was faster than AMD's tuned compiler.
The SciMark 2 micro-benchmark results were scattered among these competing Linux code compilers.