In this episode of the Let’s Talk Exascale podcast, produced by DOE’s Exascale Computing Project, the topic is PETSc—the Portable, Extensible Toolkit for Scientific Computation. It’s a suite of data structures and routines for the scalable (parallel) solution of scientific applications modeled by partial differential equations.
A team within ECP is preparing PETSc/TAO for exascale supercomputing. Richard Tran Mills and Todd Munson of Argonne National Laboratory are the project’s deputy directors. They focus on scientific collaborations and algorithms — in this podcast, they provide insights about PETSc/TAO.
PETSc supports MPI, and GPUs through Kokkos, CUDA, HIP or OpenCL, and hybrid MPI–GPU parallelism. PETSc is sometimes called PETSc/TAO because it also contains the TAO, the Toolkit for Advanced Optimization, software library.
Topics covered: PETSc/TAO’s ensemble staff and collaborators, a big-picture view of the project, strategy for accelerator support and portability, and more.
source: Scott Gibson, ECP
