EPEEC tackles issues in heterogeneous computing at PUMPS+AI Summer School 2019
OmpSs

Students and researchers recently joined the Programming and Tuning Massively Parallel Systems + Artificial Intelligence (PUMPS+AI) Summer School 2019 held at the Barcelona Supercomputing Center (BSC) on 25-29 June. 


Co-organized by EPEEC coordinator Antonio Peña, the summer school taught attendees about GPU or heterogeneous computing, one of the major concerns of the project. EPEEC´s goal is to develop and deploy a production-ready parallel programming environment that turns upcoming overwhelmingly-heterogeneous exascale supercomputers into manageable platforms for domain application developers.


Among the various courses in CUDA architecture and programming languages for parallel computing at PUMPS+AI was a lecture and tutorial of OmpSs delivered as part of the EPEEC project. OmpSs is a programming model that enables asynchronous parallelism to support heterogeneous computing. 


“The PUMPS+AI Summer School brings together a large number of researchers interested in heterogeneous computing. It is a good opportunity for the EPEEC project to show how we tackle issues in this field,” said Antonio Peña.

 
PUMPS+AI Summer School is aimed at enriching the skills of researchers, graduate students and teachers with cutting-edge technique and hands-on experience in developing applications for many-core processors with massively parallel computing resources like GPU accelerators. 


It is hosted annually by BSC in association with Universitat Politécnica de Catalunya (UPC). 


This year´s program included a hackathon and a poster session for research related to GPUs or AI.