Webinar: Getting Performance from OpenMP Programs on NUMA Architectures

Most contemporary shared memory systems expose a non-uniform memory architecture (NUMA) with implications on application performance. However, the OpenMP programming model does not provide explicit support for that. This 30-minute live webinar will discuss the approaches to getting the best performance from OpenMP applications on NUMA architecture.

The recording from the webinar is now available here: https://pop-coe.eu/blog/2nd-pop-webinar-getting-performance-from-openmp-programs-on-numa-architectures.

Leave a Reply