Evaluating CUDA Portability with HIPCL and DPCT

Zheming Jin; Jeffrey Vetter

doi:10.1109/IPDPSW52791.2021.00065

Evaluating CUDA Portability with HIPCL and DPCT

Zheming Jin, Jeffrey Vetter

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

6 Scopus citations

Abstract

HIPCL is expanding the scope of the CUDA portability route from an AMD platform to an OpenCL platform. In the meantime, the Intel DPC++ Compatibility Tool (DPCT) is migrating a CUDA program to a data parallel C++ (DPC++) program. Towards the goal of portability enhancement, we evaluate the performance of the CUDA applications from Rodinia, SHOC, and proxy applications ported using HIPCL and DPCT on Intel GPUs. After profiling the ported programs, we aim to understand their performance gaps, and optimize codes converted by DPCT to improve their performance. The open-source repository for the CUDA, HIP, and DPCT programs will be useful for the development of a translator.

Original language	English
Title of host publication	2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	371-376
Number of pages	6
ISBN (Electronic)	9781665435772
DOIs	https://doi.org/10.1109/IPDPSW52791.2021.00065
State	Published - Jun 2021
Event	2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - Virtual, Portland, United States Duration: May 17 2021 → …

Publication series

Name	2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021

Conference

Conference	2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021
Country/Territory	United States
City	Virtual, Portland
Period	05/17/21 → …

Funding

ACKNOWLEDGMENT We sincerely appreciate the reviewers for their constructive criticism and the development teams for improving HIPCL and DPCT. This research was supported by the US Department of Energy Advanced Scientific Computing Research program under Contract No. DE-AC05-00OR22725. The results presented were obtained using the Chameleon testbed and the Intel DevCloud. This manuscript has been authored by UT-Battelle, LLC, under contract DE-AC05-00OR22725 with the US Department of Energy (DOE). The US government retains and the publisher, by accepting the article for publication, acknowledges that the US government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for US government purposes. DOE will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan).

Keywords

CUDA
CUDA Portability
DPC++
HIP
OpenCL

Access to Document

10.1109/IPDPSW52791.2021.00065

Cite this

Jin, Z., & Vetter, J. (2021). Evaluating CUDA Portability with HIPCL and DPCT. In 2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021 (pp. 371-376). Article 9460636 (2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IPDPSW52791.2021.00065

Jin, Zheming ; Vetter, Jeffrey. / Evaluating CUDA Portability with HIPCL and DPCT. 2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021. Institute of Electrical and Electronics Engineers Inc., 2021. pp. 371-376 (2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021).

@inproceedings{3406170102d64fa4a7eb9f9310f548ea,

title = "Evaluating CUDA Portability with HIPCL and DPCT",

abstract = "HIPCL is expanding the scope of the CUDA portability route from an AMD platform to an OpenCL platform. In the meantime, the Intel DPC++ Compatibility Tool (DPCT) is migrating a CUDA program to a data parallel C++ (DPC++) program. Towards the goal of portability enhancement, we evaluate the performance of the CUDA applications from Rodinia, SHOC, and proxy applications ported using HIPCL and DPCT on Intel GPUs. After profiling the ported programs, we aim to understand their performance gaps, and optimize codes converted by DPCT to improve their performance. The open-source repository for the CUDA, HIP, and DPCT programs will be useful for the development of a translator.",

keywords = "CUDA, CUDA Portability, DPC++, HIP, OpenCL",

author = "Zheming Jin and Jeffrey Vetter",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 ; Conference date: 17-05-2021",

year = "2021",

month = jun,

doi = "10.1109/IPDPSW52791.2021.00065",

language = "English",

series = "2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "371--376",

booktitle = "2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021",

}

Jin, Z & Vetter, J 2021, Evaluating CUDA Portability with HIPCL and DPCT. in 2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021., 9460636, 2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021, Institute of Electrical and Electronics Engineers Inc., pp. 371-376, 2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021, Virtual, Portland, United States, 05/17/21. https://doi.org/10.1109/IPDPSW52791.2021.00065

Evaluating CUDA Portability with HIPCL and DPCT. / Jin, Zheming; Vetter, Jeffrey.
2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021. Institute of Electrical and Electronics Engineers Inc., 2021. p. 371-376 9460636 (2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Evaluating CUDA Portability with HIPCL and DPCT

AU - Jin, Zheming

AU - Vetter, Jeffrey

PY - 2021/6

Y1 - 2021/6

N2 - HIPCL is expanding the scope of the CUDA portability route from an AMD platform to an OpenCL platform. In the meantime, the Intel DPC++ Compatibility Tool (DPCT) is migrating a CUDA program to a data parallel C++ (DPC++) program. Towards the goal of portability enhancement, we evaluate the performance of the CUDA applications from Rodinia, SHOC, and proxy applications ported using HIPCL and DPCT on Intel GPUs. After profiling the ported programs, we aim to understand their performance gaps, and optimize codes converted by DPCT to improve their performance. The open-source repository for the CUDA, HIP, and DPCT programs will be useful for the development of a translator.

AB - HIPCL is expanding the scope of the CUDA portability route from an AMD platform to an OpenCL platform. In the meantime, the Intel DPC++ Compatibility Tool (DPCT) is migrating a CUDA program to a data parallel C++ (DPC++) program. Towards the goal of portability enhancement, we evaluate the performance of the CUDA applications from Rodinia, SHOC, and proxy applications ported using HIPCL and DPCT on Intel GPUs. After profiling the ported programs, we aim to understand their performance gaps, and optimize codes converted by DPCT to improve their performance. The open-source repository for the CUDA, HIP, and DPCT programs will be useful for the development of a translator.

KW - CUDA

KW - CUDA Portability

KW - DPC++

KW - HIP

KW - OpenCL

UR - http://www.scopus.com/inward/record.url?scp=85114442673&partnerID=8YFLogxK

U2 - 10.1109/IPDPSW52791.2021.00065

DO - 10.1109/IPDPSW52791.2021.00065

M3 - Conference contribution

AN - SCOPUS:85114442673

T3 - 2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021

SP - 371

EP - 376

BT - 2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021

Y2 - 17 May 2021

ER -

Jin Z, Vetter J. Evaluating CUDA Portability with HIPCL and DPCT. In 2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021. Institute of Electrical and Electronics Engineers Inc. 2021. p. 371-376. 9460636. (2021 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2021 - In conjunction with IEEE IPDPS 2021). doi: 10.1109/IPDPSW52791.2021.00065

Evaluating CUDA Portability with HIPCL and DPCT

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this