Arm SVE Hackathon

Date: 17/Dec/2020 Time: 09:00 - 18:00

Place:
Virtual event using Zoom

Target group: This hackathon targets application developers that are interested in investigating the impact of novel SVE extensions in their HPC applications.

Cost: Registration costs will be covered by the organizers

Primary tabs

Objectives

This hackathon targets application developers that are interested in investigating the impact of novel SVE extensions in their High Performance Computing (HPC) and Machine Learning (ML) applications. Arm researchers and engineers will describe the different HPC tools available from Arm (including compilers, libraries, debugging and profiling tools) to explore the benefits of SVE extensions in a particular application. Also, they will guide the attendees when porting their applications of interest to the first CPU to implement SVE, the Fujitsu A64FX. Also, representative mini-applications from well-known HPC benchmark suites will be provided to attendees that prefer to explore the benefits SVE capabilities with simpler codes. In order to ease this experience, the organizers will provide access to the CTE-ARM cluster from Fujitsu hosted at BSC facilities with all the required tools already installed.

This event is supported by the Arm-BSC Centre of Excellence and the Mont-Blanc 2020 European project.

Requirements

Requirements:
- Master in Computer Sciences or similar.
- Expert programmer in C/C++, OpenMP, MPI
- Experience as a user of an HPC cluster.
 

Academic Staff

Co-organizers of the Arm SVE Hackathon: Conrad Hillairet (Arm), Phil Ridley (Arm), Javier Setoain (Arm), Filippo Mantovani (BSC) and Miquel Moreto (BSC).

Materials

Slides, source code, and all other training materials can be browsed and downloaded at https://gitlab.com/arm-hpc/training/arm-sve-tools/. System information can be found at https://www.bsc.es/user-support/arm.php

Further information

Login details will be emailed to the address you provided in your registration. If you have not registered for the event then you will not be able to access the Fujitsu training cluster.