Task-based parallelism in scientific computing (HPC2N/PRACE, May 2021)

From SNIC Documentation
Jump to: navigation, search
Name Task-based parallelism in scientific computing (HPC2N/PRACE, May 2021)
Description Task-based parallelism in scientific computing
Type of event Course
Location Online, organized from Umeå
Start date 2021-05-10
End date 2021-05-12

Course overview

The purpose of the course is to learn when a code could benefit from task-based parallelism, and how to apply it. A task-based algorithm does not call the associated computation kernels directly, instead it is the role of a runtime system to schedule the task to various computational resources, such as CPUs and GPUs. The main benefit is that the underlying parallelism is exposed automatically. The course mainly focuses on the task-pragmas implemented in the newer incarnations of OpenMP. Other task-based runtime systems, e.g., StarPU, and GPU offloading are briefly discussed.

This course will be three half-days. It will have lectures as well as some hands-on sessions where you get to try out what you have just learned.

NOTE: Due to the COVID-19 situation, the course will be fully online, through Zoom.

This HPC2N/SNIC course is part of the PRACE Training courses. It is open for participants from PRACE member countries.

Course web page and registration

For more information and registration refer to the course webpage