Using C++ AMP to accelerate HPC applications on multiple platforms

M. Graham Lopez; Christopher Bergstrom; Ying Wai Li; Wael Elwasif; Oscar Hernandez

doi:10.1007/978-3-319-46079-6_38

Using C++ AMP to accelerate HPC applications on multiple platforms

M. Graham Lopez, Christopher Bergstrom, Ying Wai Li, Wael Elwasif, Oscar Hernandez

Programming Systems

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

1 Scopus citations

Abstract

Many high-end HPC systems support accelerators in their compute nodes to target a variety of workloads including highperformance computing simulations, big data / data analytics codes and visualization. To program both the CPU cores and attached accelerators, users now have multiple programming models available such as CUDA, OpenMP 4, OpenACC, C++14, etc., but some of these models fall short in their support for C++ on accelerators because they can have difficulty supporting advanced C++ features e.g. templating, class members, loops with iterators, lambdas, deep copy, etc. Usually, they either rely on unified memory, or the programming language is not aware of accelerators (e.g. C++14). In this paper, we explore a base-language solution called C++ Accelerated Massive Parallelism (AMP), which was developed by Microsoft and implemented by the PathScale ENZO compiler to program GPUs on a variety of HPC architectures including OpenPOWER and Intel Xeon. We report some prelminary in-progress results using C++ AMP to accelerate a matrix multiplication and quantum Monte Carlo application kernel, examining its expressiveness and performance using NVIDIA GPUs and the PathScale ENZO compiler. We hope that this preliminary report will provide a data point that will inform the functionality needed for future C++ standards to support accelerators with discrete memory spaces.

Original language	English
Title of host publication	High Performance Computing - ISC High Performance 2016 International Workshops ExaComm, E-MuCoCoS, HPC-IODC, IXPUG, IWOPH, P^3MA, VHPC, WOPSSS, Revised Selected
Editors	Bernd Mohr, Julian M. Kunkel, Michela Taufer
Publisher	Springer Verlag
Pages	563-576
Number of pages	14
ISBN (Print)	9783319460789
DOIs	https://doi.org/10.1007/978-3-319-46079-6_38
State	Published - 2016
Event	International Workshops on High Performance Computing, ISC High Performance 2016 and Workshop on 2nd International Workshop on Communication Architectures at Extreme Scale, ExaComm 2016, Workshop on Exascale Multi/Many Core Computing Systems, E-MuCoCoS 2016, HPC I/O in the Data Center, HPC-IODC 2016, Application Performance on Intel Xeon Phi – Being Prepared for KNL and Beyond, IXPUG 2016, International Workshop on OpenPOWER for HPC, IWOPH 2016, International Workshop on Performance Portable Programming Models for Accelerators, P^3MA 2016, Workshop on Virtualization in High-Performance Cloud Computing, VHPC 2016, Workshop on Performance and Scalability of Storage Systems, WOPSSS 2016 - Frankfurt, Germany Duration: Jun 19 2016 → Jun 23 2016

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9945 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Workshops on High Performance Computing, ISC High Performance 2016 and Workshop on 2nd International Workshop on Communication Architectures at Extreme Scale, ExaComm 2016, Workshop on Exascale Multi/Many Core Computing Systems, E-MuCoCoS 2016, HPC I/O in the Data Center, HPC-IODC 2016, Application Performance on Intel Xeon Phi – Being Prepared for KNL and Beyond, IXPUG 2016, International Workshop on OpenPOWER for HPC, IWOPH 2016, International Workshop on Performance Portable Programming Models for Accelerators, P^3MA 2016, Workshop on Virtualization in High-Performance Cloud Computing, VHPC 2016, Workshop on Performance and Scalability of Storage Systems, WOPSSS 2016
Country/Territory	Germany
City	Frankfurt
Period	06/19/16 → 06/23/16

Funding

This material is based upon work supported by the U.S. Department of Energy, Office of science, and this research used resources of the Oak Ridge Leadership Computing Facility at the Oak Ridge National Laboratory, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC05-00OR22725.

Keywords

Accelerator programming
C++ AMP
C++ for Accelerators
HPC

Access to Document

10.1007/978-3-319-46079-6_38

Cite this

Lopez, M. G., Bergstrom, C., Li, Y. W., Elwasif, W., & Hernandez, O. (2016). Using C++ AMP to accelerate HPC applications on multiple platforms. In B. Mohr, J. M. Kunkel, & M. Taufer (Eds.), High Performance Computing - ISC High Performance 2016 International Workshops ExaComm, E-MuCoCoS, HPC-IODC, IXPUG, IWOPH, P^3MA, VHPC, WOPSSS, Revised Selected (pp. 563-576). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9945 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-46079-6_38

Lopez, M. Graham ; Bergstrom, Christopher ; Li, Ying Wai et al. / Using C++ AMP to accelerate HPC applications on multiple platforms. High Performance Computing - ISC High Performance 2016 International Workshops ExaComm, E-MuCoCoS, HPC-IODC, IXPUG, IWOPH, P^3MA, VHPC, WOPSSS, Revised Selected. editor / Bernd Mohr ; Julian M. Kunkel ; Michela Taufer. Springer Verlag, 2016. pp. 563-576 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{1b78c61236af4ef5a952d9323777f309,

title = "Using C++ AMP to accelerate HPC applications on multiple platforms",

abstract = "Many high-end HPC systems support accelerators in their compute nodes to target a variety of workloads including highperformance computing simulations, big data / data analytics codes and visualization. To program both the CPU cores and attached accelerators, users now have multiple programming models available such as CUDA, OpenMP 4, OpenACC, C++14, etc., but some of these models fall short in their support for C++ on accelerators because they can have difficulty supporting advanced C++ features e.g. templating, class members, loops with iterators, lambdas, deep copy, etc. Usually, they either rely on unified memory, or the programming language is not aware of accelerators (e.g. C++14). In this paper, we explore a base-language solution called C++ Accelerated Massive Parallelism (AMP), which was developed by Microsoft and implemented by the PathScale ENZO compiler to program GPUs on a variety of HPC architectures including OpenPOWER and Intel Xeon. We report some prelminary in-progress results using C++ AMP to accelerate a matrix multiplication and quantum Monte Carlo application kernel, examining its expressiveness and performance using NVIDIA GPUs and the PathScale ENZO compiler. We hope that this preliminary report will provide a data point that will inform the functionality needed for future C++ standards to support accelerators with discrete memory spaces.",

keywords = "Accelerator programming, C++ AMP, C++ for Accelerators, HPC",

author = "Lopez, {M. Graham} and Christopher Bergstrom and Li, {Ying Wai} and Wael Elwasif and Oscar Hernandez",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2016.; International Workshops on High Performance Computing, ISC High Performance 2016 and Workshop on 2nd International Workshop on Communication Architectures at Extreme Scale, ExaComm 2016, Workshop on Exascale Multi/Many Core Computing Systems, E-MuCoCoS 2016, HPC I/O in the Data Center, HPC-IODC 2016, Application Performance on Intel Xeon Phi – Being Prepared for KNL and Beyond, IXPUG 2016, International Workshop on OpenPOWER for HPC, IWOPH 2016, International Workshop on Performance Portable Programming Models for Accelerators, P^3MA 2016, Workshop on Virtualization in High-Performance Cloud Computing, VHPC 2016, Workshop on Performance and Scalability of Storage Systems, WOPSSS 2016 ; Conference date: 19-06-2016 Through 23-06-2016",

year = "2016",

doi = "10.1007/978-3-319-46079-6_38",

language = "English",

isbn = "9783319460789",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "563--576",

editor = "Bernd Mohr and Kunkel, {Julian M.} and Michela Taufer",

booktitle = "High Performance Computing - ISC High Performance 2016 International Workshops ExaComm, E-MuCoCoS, HPC-IODC, IXPUG, IWOPH, P^3MA, VHPC, WOPSSS, Revised Selected",

}

Lopez, MG, Bergstrom, C, Li, YW, Elwasif, W & Hernandez, O 2016, Using C++ AMP to accelerate HPC applications on multiple platforms. in B Mohr, JM Kunkel & M Taufer (eds), High Performance Computing - ISC High Performance 2016 International Workshops ExaComm, E-MuCoCoS, HPC-IODC, IXPUG, IWOPH, P^3MA, VHPC, WOPSSS, Revised Selected. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9945 LNCS, Springer Verlag, pp. 563-576, International Workshops on High Performance Computing, ISC High Performance 2016 and Workshop on 2nd International Workshop on Communication Architectures at Extreme Scale, ExaComm 2016, Workshop on Exascale Multi/Many Core Computing Systems, E-MuCoCoS 2016, HPC I/O in the Data Center, HPC-IODC 2016, Application Performance on Intel Xeon Phi – Being Prepared for KNL and Beyond, IXPUG 2016, International Workshop on OpenPOWER for HPC, IWOPH 2016, International Workshop on Performance Portable Programming Models for Accelerators, P^3MA 2016, Workshop on Virtualization in High-Performance Cloud Computing, VHPC 2016, Workshop on Performance and Scalability of Storage Systems, WOPSSS 2016, Frankfurt, Germany, 06/19/16. https://doi.org/10.1007/978-3-319-46079-6_38

Using C++ AMP to accelerate HPC applications on multiple platforms. / Lopez, M. Graham; Bergstrom, Christopher; Li, Ying Wai et al.
High Performance Computing - ISC High Performance 2016 International Workshops ExaComm, E-MuCoCoS, HPC-IODC, IXPUG, IWOPH, P^3MA, VHPC, WOPSSS, Revised Selected. ed. / Bernd Mohr; Julian M. Kunkel; Michela Taufer. Springer Verlag, 2016. p. 563-576 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9945 LNCS).

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

TY - GEN

T1 - Using C++ AMP to accelerate HPC applications on multiple platforms

AU - Lopez, M. Graham

AU - Bergstrom, Christopher

AU - Li, Ying Wai

AU - Elwasif, Wael

AU - Hernandez, Oscar

N1 - Publisher Copyright: © Springer International Publishing AG 2016.

PY - 2016

Y1 - 2016

N2 - Many high-end HPC systems support accelerators in their compute nodes to target a variety of workloads including highperformance computing simulations, big data / data analytics codes and visualization. To program both the CPU cores and attached accelerators, users now have multiple programming models available such as CUDA, OpenMP 4, OpenACC, C++14, etc., but some of these models fall short in their support for C++ on accelerators because they can have difficulty supporting advanced C++ features e.g. templating, class members, loops with iterators, lambdas, deep copy, etc. Usually, they either rely on unified memory, or the programming language is not aware of accelerators (e.g. C++14). In this paper, we explore a base-language solution called C++ Accelerated Massive Parallelism (AMP), which was developed by Microsoft and implemented by the PathScale ENZO compiler to program GPUs on a variety of HPC architectures including OpenPOWER and Intel Xeon. We report some prelminary in-progress results using C++ AMP to accelerate a matrix multiplication and quantum Monte Carlo application kernel, examining its expressiveness and performance using NVIDIA GPUs and the PathScale ENZO compiler. We hope that this preliminary report will provide a data point that will inform the functionality needed for future C++ standards to support accelerators with discrete memory spaces.

AB - Many high-end HPC systems support accelerators in their compute nodes to target a variety of workloads including highperformance computing simulations, big data / data analytics codes and visualization. To program both the CPU cores and attached accelerators, users now have multiple programming models available such as CUDA, OpenMP 4, OpenACC, C++14, etc., but some of these models fall short in their support for C++ on accelerators because they can have difficulty supporting advanced C++ features e.g. templating, class members, loops with iterators, lambdas, deep copy, etc. Usually, they either rely on unified memory, or the programming language is not aware of accelerators (e.g. C++14). In this paper, we explore a base-language solution called C++ Accelerated Massive Parallelism (AMP), which was developed by Microsoft and implemented by the PathScale ENZO compiler to program GPUs on a variety of HPC architectures including OpenPOWER and Intel Xeon. We report some prelminary in-progress results using C++ AMP to accelerate a matrix multiplication and quantum Monte Carlo application kernel, examining its expressiveness and performance using NVIDIA GPUs and the PathScale ENZO compiler. We hope that this preliminary report will provide a data point that will inform the functionality needed for future C++ standards to support accelerators with discrete memory spaces.

KW - Accelerator programming

KW - C++ AMP

KW - C++ for Accelerators

KW - HPC

UR - http://www.scopus.com/inward/record.url?scp=84992616431&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-46079-6_38

DO - 10.1007/978-3-319-46079-6_38

M3 - Conference contribution

AN - SCOPUS:84992616431

SN - 9783319460789

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 563

EP - 576

BT - High Performance Computing - ISC High Performance 2016 International Workshops ExaComm, E-MuCoCoS, HPC-IODC, IXPUG, IWOPH, P^3MA, VHPC, WOPSSS, Revised Selected

A2 - Mohr, Bernd

A2 - Kunkel, Julian M.

A2 - Taufer, Michela

PB - Springer Verlag

T2 - International Workshops on High Performance Computing, ISC High Performance 2016 and Workshop on 2nd International Workshop on Communication Architectures at Extreme Scale, ExaComm 2016, Workshop on Exascale Multi/Many Core Computing Systems, E-MuCoCoS 2016, HPC I/O in the Data Center, HPC-IODC 2016, Application Performance on Intel Xeon Phi – Being Prepared for KNL and Beyond, IXPUG 2016, International Workshop on OpenPOWER for HPC, IWOPH 2016, International Workshop on Performance Portable Programming Models for Accelerators, P^3MA 2016, Workshop on Virtualization in High-Performance Cloud Computing, VHPC 2016, Workshop on Performance and Scalability of Storage Systems, WOPSSS 2016

Y2 - 19 June 2016 through 23 June 2016

ER -

Lopez MG, Bergstrom C, Li YW, Elwasif W , Hernandez O. Using C++ AMP to accelerate HPC applications on multiple platforms. In Mohr B, Kunkel JM, Taufer M, editors, High Performance Computing - ISC High Performance 2016 International Workshops ExaComm, E-MuCoCoS, HPC-IODC, IXPUG, IWOPH, P^3MA, VHPC, WOPSSS, Revised Selected. Springer Verlag. 2016. p. 563-576. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-46079-6_38

Using C++ AMP to accelerate HPC applications on multiple platforms

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this