PGI (The Portland Group) is working with the NNSA to create an open-source Fortran compiler designed to integrate with the widely used LLVM compiler infrastructure. Recently, PGI announced comparable x86 multicore OpenACC performance as compared the the Intel compiler using OpenMP. This move ensure that PGI will have access to the latest and best information about the AVX-512 instruction set and optimization technology for Intel Xeon Phi devices. According to Lawrence Livermore Laboratory, source code for the resulting Fortran front-end is expected to be first available in open-source form in late 2016. Early reaction from the HPC community has been immediate and very enthusiastic.
Along with Intel Xeon Phi, this move also puts PGI in an excellent position with IBM and OpenPower. Again according to the LLNL release:
“We have been working with the LLVM community for more than a year on the development of parallelizing C and C++ compilers for heterogeneous OpenPOWER HPC systems with closely coupled GPU accelerators,” said Dave Turek, vice president of Exascale Computing at IBM. “LLVM-based compilers provide an important addition to the OpenPOWER ecosystem joining the existing IBM XL and PGI compilers. This announcement helps to extend our current work to encompass a fully open-source Fortran compiler for these systems.”
For more information, see the LLNL press release.