Treffer: Static Graphs for Coding Productivity in OpenACC

Title:
Static Graphs for Coding Productivity in OpenACC
Authors:
Toledo, Leonel, Valero Lara, Pedro, Vetter, Jeffrey, Peña, Antonio
Contributors:
Barcelona Supercomputing Center
Publisher Information:
Institute of Electrical and Electronics Engineers (IEEE)
Publication Year:
2022
Collection:
Universitat Politècnica de Catalunya, BarcelonaTech: UPCommons - Global access to UPC knowledge
Subject Terms:
Àrees temàtiques de la UPC::Informàtica::Enginyeria del software, GPUs (Graphics processing units), C++/CLI (Computer program language), C (Computer program language), FORTRAN (Computer program language), Coding Productivity, Tasking, Data Dependencies, Static Graph, OpenACC, Particle Swarm Optimization, Supercomputadors
Document Type:
Konferenz conference object
File Description:
6 p.; application/pdf
Language:
English
Relation:
https://ieeexplore.ieee.org/document/9680348; info:eu-repo/grantAgreement/EC/H2020/801051/EU/European joint Effort toward a Highly Productive Programming Environment for Heterogeneous Exascale Computing (EPEEC)/EPEEC; http://hdl.handle.net/2117/363953
DOI:
10.1109/HiPC53243.2021.00050
Availability:
http://hdl.handle.net/2117/363953
https://doi.org/10.1109/HiPC53243.2021.00050
Rights:
Open Access
Accession Number:
edsbas.A28C3570
Database:
BASE

Weitere Informationen

The main contribution of this work is to increase the coding productivity for GPU programming by using the concept of Static Graphs. To do so, we have combined the new CUDA Graph API with the OpenACC programming model. We use as test cases a well-known and widely used problems in HPC and AI: the Particle Swarm Optimization. We complement the OpenACC functionality with the use of CUDA Graph, achieving accelerations of more than one order of magnitude, and a performance very close to a reference and optimized CUDA code. Finally, we propose a new specification to incorporate the concept of Static Graphs into the OpenACC specification. ; This project has received funding from the EPEEC project from the European Union’s Horizon 2020 Research and Innovation program under grant agreement No. 801051. ; Peer Reviewed ; Postprint (author's final draft)