For the FFTE test profile, there is not a huge change in performance between the eleven different compilers that were tested on this Intel Core i7 Ivy Bridge processor.
When running the HMMer Pfam test, the fastest binaries produced were built under GCC. GCC 4.6, 4.7, and 4.8 were all running at around the same speed with the corei7-avx optimizations. LLVM/Clang and DragonEgg were slower although DragonEgg from SVN with LLVM 3.1 from Git had caught up to nearly the speed of GCC when applying both the LLVM and GCC optimizers.
The fastest MAFFT multi-sequence results were on Open64 5.0 although the GCC results were right behind and then behind that were all of the LLVM/Clang/DragonEgg-built binaries.