Hardware Acceleration of High-Performance Computational Flow Dynamics Using High-Bandwidth Memory-Enabled Field-Programmable Gate Arrays
Scientific computing is at the core of many High-Performance Computing applications, including computational flow dynamics. Because of the utmost importance to simulate increasingly larger computational models, hardware acceleration is receiving increased attention due to its potential to maximize t...
Published in: | ACM Transactions on Reconfigurable Technology and Systems |
---|---|
Main Authors: | , , , , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Association for Computing Machinery (ACM)
2021
|
Subjects: | |
Online Access: | http://dx.doi.org/10.1145/3476229 https://dl.acm.org/doi/pdf/10.1145/3476229 |
id |
cracm:10.1145/3476229 |
---|---|
record_format |
openpolar |
spelling |
cracm:10.1145/3476229 2024-05-19T07:44:24+00:00 Hardware Acceleration of High-Performance Computational Flow Dynamics Using High-Bandwidth Memory-Enabled Field-Programmable Gate Arrays Hogervorst, Tom Nane, Răzvan Marchiori, Giacomo Qiu, Tong Dong Blatt, Markus Rustad, Alf Birger 2021 http://dx.doi.org/10.1145/3476229 https://dl.acm.org/doi/pdf/10.1145/3476229 en eng Association for Computing Machinery (ACM) ACM Transactions on Reconfigurable Technology and Systems volume 15, issue 2, page 1-35 ISSN 1936-7406 1936-7414 journal-article 2021 cracm https://doi.org/10.1145/3476229 2024-05-01T06:44:59Z Scientific computing is at the core of many High-Performance Computing applications, including computational flow dynamics. Because of the utmost importance to simulate increasingly larger computational models, hardware acceleration is receiving increased attention due to its potential to maximize the performance of scientific computing. Field-Programmable Gate Arrays could accelerate scientific computing because of the possibility to fully customize the memory hierarchy important in irregular applications such as iterative linear solvers. In this article, we study the potential of using Field-Programmable Gate Arrays in High-Performance Computing because of the rapid advances in reconfigurable hardware, such as the increase in on-chip memory size, increasing number of logic cells, and the integration of High-Bandwidth Memories on board. To perform this study, we propose a novel Sparse Matrix-Vector multiplication unit and an ILU0 preconditioner tightly integrated with a BiCGStab solver kernel. We integrate the developed preconditioned iterative solver in Flow from the Open Porous Media project, a state-of-the-art open source reservoir simulator. Finally, we perform a thorough evaluation of the FPGA solver kernel in both stand-alone mode and integrated in the reservoir simulator, using the NORNE field, a real-world case reservoir model using a grid with more than 10 5 cells and using three unknowns per cell. Article in Journal/Newspaper Norne field ACM Publications (Association for Computing Machinery) ACM Transactions on Reconfigurable Technology and Systems 15 2 1 35 |
institution |
Open Polar |
collection |
ACM Publications (Association for Computing Machinery) |
op_collection_id |
cracm |
language |
English |
description |
Scientific computing is at the core of many High-Performance Computing applications, including computational flow dynamics. Because of the utmost importance to simulate increasingly larger computational models, hardware acceleration is receiving increased attention due to its potential to maximize the performance of scientific computing. Field-Programmable Gate Arrays could accelerate scientific computing because of the possibility to fully customize the memory hierarchy important in irregular applications such as iterative linear solvers. In this article, we study the potential of using Field-Programmable Gate Arrays in High-Performance Computing because of the rapid advances in reconfigurable hardware, such as the increase in on-chip memory size, increasing number of logic cells, and the integration of High-Bandwidth Memories on board. To perform this study, we propose a novel Sparse Matrix-Vector multiplication unit and an ILU0 preconditioner tightly integrated with a BiCGStab solver kernel. We integrate the developed preconditioned iterative solver in Flow from the Open Porous Media project, a state-of-the-art open source reservoir simulator. Finally, we perform a thorough evaluation of the FPGA solver kernel in both stand-alone mode and integrated in the reservoir simulator, using the NORNE field, a real-world case reservoir model using a grid with more than 10 5 cells and using three unknowns per cell. |
format |
Article in Journal/Newspaper |
author |
Hogervorst, Tom Nane, Răzvan Marchiori, Giacomo Qiu, Tong Dong Blatt, Markus Rustad, Alf Birger |
spellingShingle |
Hogervorst, Tom Nane, Răzvan Marchiori, Giacomo Qiu, Tong Dong Blatt, Markus Rustad, Alf Birger Hardware Acceleration of High-Performance Computational Flow Dynamics Using High-Bandwidth Memory-Enabled Field-Programmable Gate Arrays |
author_facet |
Hogervorst, Tom Nane, Răzvan Marchiori, Giacomo Qiu, Tong Dong Blatt, Markus Rustad, Alf Birger |
author_sort |
Hogervorst, Tom |
title |
Hardware Acceleration of High-Performance Computational Flow Dynamics Using High-Bandwidth Memory-Enabled Field-Programmable Gate Arrays |
title_short |
Hardware Acceleration of High-Performance Computational Flow Dynamics Using High-Bandwidth Memory-Enabled Field-Programmable Gate Arrays |
title_full |
Hardware Acceleration of High-Performance Computational Flow Dynamics Using High-Bandwidth Memory-Enabled Field-Programmable Gate Arrays |
title_fullStr |
Hardware Acceleration of High-Performance Computational Flow Dynamics Using High-Bandwidth Memory-Enabled Field-Programmable Gate Arrays |
title_full_unstemmed |
Hardware Acceleration of High-Performance Computational Flow Dynamics Using High-Bandwidth Memory-Enabled Field-Programmable Gate Arrays |
title_sort |
hardware acceleration of high-performance computational flow dynamics using high-bandwidth memory-enabled field-programmable gate arrays |
publisher |
Association for Computing Machinery (ACM) |
publishDate |
2021 |
url |
http://dx.doi.org/10.1145/3476229 https://dl.acm.org/doi/pdf/10.1145/3476229 |
genre |
Norne field |
genre_facet |
Norne field |
op_source |
ACM Transactions on Reconfigurable Technology and Systems volume 15, issue 2, page 1-35 ISSN 1936-7406 1936-7414 |
op_doi |
https://doi.org/10.1145/3476229 |
container_title |
ACM Transactions on Reconfigurable Technology and Systems |
container_volume |
15 |
container_issue |
2 |
container_start_page |
1 |
op_container_end_page |
35 |
_version_ |
1799484185157566464 |