In this video from FOSDEM’19, Thomas Schwinge from Mentor presents: Speeding up Programs with OpenACC in GCC.
Proven in production use for decades, GCC (the GNU Compiler Collection) offers C, C++, Fortran, and other compilers for a multitude of target systems. Over the last few years, we — formerly known as “CodeSourcery”, now a group in “Mentor, a Siemens Business” — added support for the directive-based OpenACC programming model. Requiring only few changes to your existing source code, OpenACC allows for easy parallelization and code offloading to accelerators such as GPUs. We will present a short introduction of GCC and OpenACC, implementation status, examples, and performance results.
OpenACC is a user-driven directive-based performance-portable parallel programming model designed for scientists and engineers interested in porting their codes to a wide-variety of heterogeneous HPC hardware platforms and architectures with significantly less programming effort than required with a low-level model.”
Thomas Schwinge is a long-time Free Software/Open Source software user and contributor — and FOSDEM visitor. He holds a degree in Electrical Engineering and Computer Engineering from the University of Stuttgart. For several years now, he has primarily been working on GCC (GNU Compiler Collection), in particular adding support for OpenACC (a directive-based parallel programming model) with code offloading to Nvidia GPUs, employed by Mentor, a Siemens Business (… what used to be Mentor Graphics, what used to be CodeSourcery).