Evaluating the performance of HPC-style SYCL applications

Tom Deakin; Simon McIntosh-Smith

doi:10.1145/3388333.3388643

Evaluating the performance of HPC-style SYCL applications

Tom Deakin^*, Simon McIntosh-Smith^*

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

1 Citation (Scopus)

1648 Downloads (Pure)

Abstract

SYCL is a parallel programming model for developing single-source programs for running on heterogeneous platforms. To this end, it allows for one code to be written which can run on a different architectures. For this study, we develop applications in SYCL which are representative of those often used in High-Performance Computing. Their performance is benchmarked on a variety of CPU and GPU architectures from multiple vendors, and compared to well optimised versions written in OpenCL and other parallel programming models.

Original language	English
Title of host publication	IWOCL '20
Subtitle of host publication	Proceedings of the International Workshop on OpenCL
Publisher	Association for Computing Machinery (ACM)
Number of pages	11
ISBN (Electronic)	978-1-4503-7531-3
DOIs	https://doi.org/10.1145/3388333.3388643
Publication status	Published - 1 Apr 2020
Event	International Workshop on OpenCL - Leibniz Supercomputing Centre, Munich, Germany Duration: 27 Apr 2020 → 29 Apr 2020

Conference

Conference	International Workshop on OpenCL
Abbreviated title	IWOCL
Country/Territory	Germany
City	Munich
Period	27/04/20 → 29/04/20

Keywords

SYCL
GPGPUs
performance portability
benchmarking

Access to Document

10.1145/3388333.3388643

Full-text PDF (accepted author manuscript)
This is the author accepted manuscript (AAM). The final published version (version of record) is available online via the Association for Computing Machinery at https://dl.acm.org/doi/10.1145/3388333.3388643 . Please refer to any applicable terms of use of the publisher.
Accepted author manuscript, 465 KB

Handle.net

Persistent link

Cite this

@inproceedings{910a7be5f09e40eaa113c1a86e63040c,

title = "Evaluating the performance of HPC-style SYCL applications",

abstract = "SYCL is a parallel programming model for developing single-source programs for running on heterogeneous platforms. To this end, it allows for one code to be written which can run on a different architectures. For this study, we develop applications in SYCL which are representative of those often used in High-Performance Computing. Their performance is benchmarked on a variety of CPU and GPU architectures from multiple vendors, and compared to well optimised versions written in OpenCL and other parallel programming models.",

keywords = "SYCL, GPGPUs, performance portability, benchmarking",

author = "Tom Deakin and Simon McIntosh-Smith",

year = "2020",

month = apr,

day = "1",

doi = "10.1145/3388333.3388643",

language = "English",

booktitle = "IWOCL '20",

publisher = "Association for Computing Machinery (ACM)",

address = "United States",

note = "International Workshop on OpenCL, IWOCL ; Conference date: 27-04-2020 Through 29-04-2020",

}

TY - GEN

T1 - Evaluating the performance of HPC-style SYCL applications

AU - Deakin, Tom

AU - McIntosh-Smith, Simon

PY - 2020/4/1

Y1 - 2020/4/1

N2 - SYCL is a parallel programming model for developing single-source programs for running on heterogeneous platforms. To this end, it allows for one code to be written which can run on a different architectures. For this study, we develop applications in SYCL which are representative of those often used in High-Performance Computing. Their performance is benchmarked on a variety of CPU and GPU architectures from multiple vendors, and compared to well optimised versions written in OpenCL and other parallel programming models.

AB - SYCL is a parallel programming model for developing single-source programs for running on heterogeneous platforms. To this end, it allows for one code to be written which can run on a different architectures. For this study, we develop applications in SYCL which are representative of those often used in High-Performance Computing. Their performance is benchmarked on a variety of CPU and GPU architectures from multiple vendors, and compared to well optimised versions written in OpenCL and other parallel programming models.

KW - SYCL

KW - GPGPUs

KW - performance portability

KW - benchmarking

U2 - 10.1145/3388333.3388643

DO - 10.1145/3388333.3388643

M3 - Conference Contribution (Conference Proceeding)

BT - IWOCL '20

PB - Association for Computing Machinery (ACM)

T2 - International Workshop on OpenCL

Y2 - 27 April 2020 through 29 April 2020

ER -

Evaluating the performance of HPC-style SYCL applications

Abstract

Conference

Keywords

Access to Document

Handle.net

Fingerprint

Cite this