Podcast with Pavan Balaji and Ken Raffenetti of Argonne National Laboratory about MPI

13 Jul 2018 Argonne - The US Exascale project published "Influencing the Evolution of the MPI Standard for Optimal Exascale Scientific Applications" with Pavan Balaji and Ken Raffenetti of Argonne National Laboratory. They discuss the efforts to help MPI, the de facto programming model for parallel computing, run as efficiently as possible on exascale systems. In their opnion. MPI itself as well as the Exascale MPI project still have a long way to go.

One of their conclusions is: "New architectures are another big thing. Some of the architectures we are working on are prototypes of what we think might come, but there are some quite disruptive architectures that people are already thinking of, like, for example, memory systems with semi-reliability, that don't have the full level of reliability."

If the supercomputing centre has to use the software, it must be hardened and should be validated and verified. So the authors have already started kind of looking, with their supercomputing centre partners, to have some sort of better validation/verification for their software.

The Exascale Project website contains a transcript of the podcast.

Ad Emmen