Port an HPC Application to a GPU Using Intel oneAPI

oneAPI is a cross-industry, open, standards-based unified programming model that gives a common developer experience across accelerator architectures. It provides faster application performance, more productivity, and greater innovation. Intel® oneAPI toolkits include implementations of the oneAPI specification and complementary tools to develop and deploy applications and solutions across Intel® CPU and XPU architectures.

This training demonstrates using Intel® oneAPI tools for a basic CPU-enabled HPC application. Get strategies for:

  • Selecting potential code regions for offload to GPUs
  • Making necessary code changes and deploying the code in a heterogeneous XPU environment

Get demos of:

  • Optimizations to increase performance
  • This presentation ends with a demo on a Linux*-based machine.

 

Speaker

Dr. Amarpal Kapoor serves as the lead technical consulting engineer for cluster tools in Intel oneAPI. He joined Intel in 2016 and his principal focus area is HPC and related topics. He drives and participates in several strategic codesign, development, and performance optimization projects. He completed his PhD in computational fluid dynamics and a master's degree in computer modeling as a Zienkiewicz scholar from Swansea University, UK.