Academic Journal

Static Graphs for Coding Productivity in OpenACC

Λεπτομέρειες βιβλιογραφικής εγγραφής
Τίτλος: Static Graphs for Coding Productivity in OpenACC
Συγγραφείς: Toledo, Leonel, Valero Lara, Pedro, Vetter, Jeffrey, Peña, Antonio
Πηγή: UPCommons. Portal del coneixement obert de la UPC
Universitat Politècnica de Catalunya (UPC)
Institute of Electrical and Electronics Engineers (IEEE)
2021 IEEE 28th International Conference on High Performance Computing, Data, and Analytics (HiPC)
Στοιχεία εκδότη: IEEE, 2021.
Έτος έκδοσης: 2021
Θεματικοί όροι: Àrees temàtiques de la UPC::Informàtica::Enginyeria del software, Coding Productivity, Informàtica::Enginyeria del software [Àrees temàtiques de la UPC], Data Dependencies, 02 engineering and technology, OpenACC, C++/CLI (Computer program language), FORTRAN (Computer program language), Supercomputadors, C (Computer program language), Particle Swarm Optimization, 0202 electrical engineering, electronic engineering, information engineering, Static Graph, GPUs (Graphics processing units), Tasking
Περιγραφή: 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.
Τύπος εγγράφου: Article
Conference object
Περιγραφή αρχείου: application/pdf
DOI: 10.1109/hipc53243.2021.00050
DOI: 10.13039/100010661
Rights: IEEE Copyright
Αριθμός Καταχώρησης: edsair.doi.dedup.....c6e0db60de62e6a947f7d3c99a104098
Βάση Δεδομένων: OpenAIRE