Performance portability An implementers Perspective Versatility of HPE Cray Programming Environment CPE Platforms from HPECray Supported by CPE 2 AMD CPUs and NVIDIA GPUs AMD CPUs and AMD GPUs ID: 999199
Download Presentation The PPT/PDF document "Barbara Chapman October 13, 2022" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
1. Barbara ChapmanOctober 13, 2022Performance portability: An implementer’s Perspective
2. Versatility of HPE Cray Programming Environment (CPE)Platforms from HPE/CraySupported by CPE2AMD CPUs and NVIDIA GPUsAMD CPUs and AMD GPUsIntegrated CPU and GPU from AMDIntel CPUsIntel CPUs and NVIDIA GPUsARM CPUsIntegrated ARM CPUs and NVIDIA GPUs
3. CPE TODAYPortability via adherence to HPC-relevant standardsActive engagement in standards organizationsSupports:Fortran, C/C++MPI; OpenMP, OpenACC; OpenSHMEM; UPCEnables popular frameworks (e.g. Kokkos)3Tom Tabor presents HPCWire Editor’s Choice Award for OpenACC to Jay Gould (Cray) and Michael Wolfe (PGI)GPU(s)Host Device(CPU Multicore)One or more devices attachedAttached accelerator(s)
4. IESP Programming ModelsInternational Exascale Software ProjectProposed timelineInteroperability among existing programming modelsFault-tolerant MPIStandard programming model for heterogeneous nodes System-wide high-level programming model Exascale programming models implementedExascale programming model(s) adopted2010201120122013201420152016201720182019Your MetricCandidate exascale programming models definedwww.exascale.org