dc.contributor.author | Holm, Håvard Heitlo | |
dc.contributor.author | Brodtkorb, André R. | |
dc.contributor.author | Sætra, Martin Lilleeng | |
dc.date.accessioned | 2021-01-15T09:53:24Z | |
dc.date.accessioned | 2021-02-25T11:49:29Z | |
dc.date.available | 2021-01-15T09:53:24Z | |
dc.date.available | 2021-02-25T11:49:29Z | |
dc.date.issued | 2020 | |
dc.identifier.citation | Holm HH, Brodtkorb A, Sætra ML. Performance and Energy Efficiency of CUDA and OpenCL for GPU Computing using Python. Advances in Parallel Computing. 2020;36:593-604 | en |
dc.identifier.issn | 0927-5452 | |
dc.identifier.issn | 1879-808X | |
dc.identifier.uri | https://hdl.handle.net/10642/9742 | |
dc.description.abstract | In this work, we examine the performance and energy efficiency when using Python for developing HPC codes running on the GPU. We investigate the portability of performance and energy efficiency between CUDA and OpenCL; between GPU generations; and between low-end, mid-range and high-end GPUs. Our findings show that for some combinations of GPU and GPU code, there is a significant speedup for CUDA over OpenCL, but that this does not hold in general. Our experiments show that performance in general varies more between different GPUs, than between using CUDA and OpenCL. Finally, we show that tuning for performance is a good way of tuning for energy efficiency. | en |
dc.description.sponsorship | This work is supported by the Research Council of Norway through grant number 250935 (GPU Ocean). The GPU Ocean project received support from UNINETT Sigma2 - the National Infrastructure for High Performance Computing and Data Storage in Norway under project number nn9550k. | en |
dc.language.iso | en | en |
dc.publisher | IOS Press | en |
dc.relation.ispartofseries | Advances in Parallel Computing;Volume 36: Parallel Computing: Technology Trends | |
dc.rights | Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) License | en |
dc.subject | GPU Computing | en |
dc.subject | Compute unified device architecture | en |
dc.subject | Open compute languages | en |
dc.subject | High performance computing | en |
dc.subject | Shallow-water simulations | en |
dc.subject | Power efficiency | en |
dc.subject | OpenCL | en |
dc.title | Performance and Energy Efficiency of CUDA and OpenCL for GPU Computing using Python | en |
dc.type | Journal article | en |
dc.type | Peer reviewed | en |
dc.date.updated | 2021-01-15T09:53:23Z | |
dc.description.version | publishedVersion | en |
dc.identifier.doi | https://doi.org/10.3233/APC200089 | |
dc.identifier.cristin | 1776313 | |
dc.source.journal | Advances in Parallel Computing | |
dc.relation.projectID | Norges forskningsråd: 250935 | |
dc.relation.projectID | Norges forskningsråd: 250935 (GPU Ocean) | |
dc.relation.projectID | Notur/NorStore: NN9550K | |