Evaluation of a floating-point intensive kernel on FPGA: A case study of geodesic distance kernel

Zheming Jin; Hal Finkel; Kazutomo Yoshii; Franck Cappello

doi:10.1007/978-3-319-75178-8_53

Evaluation of a floating-point intensive kernel on FPGA: A case study of geodesic distance kernel

Zheming Jin, Hal Finkel, Kazutomo Yoshii, Franck Cappello

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

7 Scopus citations

Abstract

Heterogeneous platforms provide a promising solution for high-performance and energy-efficient computing applications. This paper presents our research on usage of heterogeneous platform for a floating-point intensive kernel. We first introduce the floating-point intensive kernel from the geographical information system. Then we analyze the FPGA designs generated by the Intel FPGA SDK for OpenCL, and evaluate the kernel performance and the floating-point error rate of the FPGA designs. Finally, we compare the performance and energy efficiency of the kernel implementations on the Arria 10 FPGA, Intel’s Xeon Phi Knights Landing CPU, and NVIDIA’s Kepler GPU. Our evaluation shows the energy efficiency of the single-precision kernel on the FPGA is 1.35X better than on the CPU and the GPU, while the energy efficiency of the double-precision kernel on the FPGA is 1.36X and 1.72X less than the CPU and GPU, respectively.

Original language	English
Title of host publication	Euro-Par 2017
Subtitle of host publication	Parallel Processing Workshops - Euro-Par 2017 International Workshops
Editors	Dora B. Heras, Luc Bouge
Publisher	Springer Verlag
Pages	664-675
Number of pages	12
ISBN (Print)	9783319751771
DOIs	https://doi.org/10.1007/978-3-319-75178-8_53
State	Published - 2018
Externally published	Yes
Event	International Workshops on Parallel Processing, Euro-Par 2017 - Santiago de Compostela, Spain Duration: Aug 28 2017 → Aug 29 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10659 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Workshops on Parallel Processing, Euro-Par 2017
Country/Territory	Spain
City	Santiago de Compostela
Period	08/28/17 → 08/29/17

Funding

Acknowledgement. We thank the anonymous reviewers and the shepherd for their comments. This research used resources of the Argonne Leadership Computing Facility, which is a DOE Office of Science User Facility supported under Contract DE-AC02-06CH11357.

Keywords

FPGA
Floating-point operation
HPC
OpenCL

Access to Document

10.1007/978-3-319-75178-8_53

Cite this

Jin, Z., Finkel, H., Yoshii, K., & Cappello, F. (2018). Evaluation of a floating-point intensive kernel on FPGA: A case study of geodesic distance kernel. In D. B. Heras, & L. Bouge (Eds.), Euro-Par 2017: Parallel Processing Workshops - Euro-Par 2017 International Workshops (pp. 664-675). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10659 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-75178-8_53

Jin, Zheming ; Finkel, Hal ; Yoshii, Kazutomo et al. / Evaluation of a floating-point intensive kernel on FPGA : A case study of geodesic distance kernel. Euro-Par 2017: Parallel Processing Workshops - Euro-Par 2017 International Workshops. editor / Dora B. Heras ; Luc Bouge. Springer Verlag, 2018. pp. 664-675 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{649b497daa2141f4be78655cd420f464,

title = "Evaluation of a floating-point intensive kernel on FPGA: A case study of geodesic distance kernel",

abstract = "Heterogeneous platforms provide a promising solution for high-performance and energy-efficient computing applications. This paper presents our research on usage of heterogeneous platform for a floating-point intensive kernel. We first introduce the floating-point intensive kernel from the geographical information system. Then we analyze the FPGA designs generated by the Intel FPGA SDK for OpenCL, and evaluate the kernel performance and the floating-point error rate of the FPGA designs. Finally, we compare the performance and energy efficiency of the kernel implementations on the Arria 10 FPGA, Intel{\textquoteright}s Xeon Phi Knights Landing CPU, and NVIDIA{\textquoteright}s Kepler GPU. Our evaluation shows the energy efficiency of the single-precision kernel on the FPGA is 1.35X better than on the CPU and the GPU, while the energy efficiency of the double-precision kernel on the FPGA is 1.36X and 1.72X less than the CPU and GPU, respectively.",

keywords = "FPGA, Floating-point operation, HPC, OpenCL",

author = "Zheming Jin and Hal Finkel and Kazutomo Yoshii and Franck Cappello",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG, part of Springer Nature 2018.; International Workshops on Parallel Processing, Euro-Par 2017 ; Conference date: 28-08-2017 Through 29-08-2017",

year = "2018",

doi = "10.1007/978-3-319-75178-8\_53",

language = "English",

isbn = "9783319751771",

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

publisher = "Springer Verlag",

pages = "664--675",

editor = "Heras, \{Dora B.\} and Luc Bouge",

booktitle = "Euro-Par 2017",

}

Jin, Z, Finkel, H, Yoshii, K & Cappello, F 2018, Evaluation of a floating-point intensive kernel on FPGA: A case study of geodesic distance kernel. in DB Heras & L Bouge (eds), Euro-Par 2017: Parallel Processing Workshops - Euro-Par 2017 International Workshops. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10659 LNCS, Springer Verlag, pp. 664-675, International Workshops on Parallel Processing, Euro-Par 2017, Santiago de Compostela, Spain, 08/28/17. https://doi.org/10.1007/978-3-319-75178-8_53

Evaluation of a floating-point intensive kernel on FPGA: A case study of geodesic distance kernel. / Jin, Zheming; Finkel, Hal; Yoshii, Kazutomo et al.
Euro-Par 2017: Parallel Processing Workshops - Euro-Par 2017 International Workshops. ed. / Dora B. Heras; Luc Bouge. Springer Verlag, 2018. p. 664-675 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10659 LNCS).

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

TY - GEN

T1 - Evaluation of a floating-point intensive kernel on FPGA

T2 - International Workshops on Parallel Processing, Euro-Par 2017

AU - Jin, Zheming

AU - Finkel, Hal

AU - Yoshii, Kazutomo

AU - Cappello, Franck

N1 - Publisher Copyright: © Springer International Publishing AG, part of Springer Nature 2018.

PY - 2018

Y1 - 2018

N2 - Heterogeneous platforms provide a promising solution for high-performance and energy-efficient computing applications. This paper presents our research on usage of heterogeneous platform for a floating-point intensive kernel. We first introduce the floating-point intensive kernel from the geographical information system. Then we analyze the FPGA designs generated by the Intel FPGA SDK for OpenCL, and evaluate the kernel performance and the floating-point error rate of the FPGA designs. Finally, we compare the performance and energy efficiency of the kernel implementations on the Arria 10 FPGA, Intel’s Xeon Phi Knights Landing CPU, and NVIDIA’s Kepler GPU. Our evaluation shows the energy efficiency of the single-precision kernel on the FPGA is 1.35X better than on the CPU and the GPU, while the energy efficiency of the double-precision kernel on the FPGA is 1.36X and 1.72X less than the CPU and GPU, respectively.

AB - Heterogeneous platforms provide a promising solution for high-performance and energy-efficient computing applications. This paper presents our research on usage of heterogeneous platform for a floating-point intensive kernel. We first introduce the floating-point intensive kernel from the geographical information system. Then we analyze the FPGA designs generated by the Intel FPGA SDK for OpenCL, and evaluate the kernel performance and the floating-point error rate of the FPGA designs. Finally, we compare the performance and energy efficiency of the kernel implementations on the Arria 10 FPGA, Intel’s Xeon Phi Knights Landing CPU, and NVIDIA’s Kepler GPU. Our evaluation shows the energy efficiency of the single-precision kernel on the FPGA is 1.35X better than on the CPU and the GPU, while the energy efficiency of the double-precision kernel on the FPGA is 1.36X and 1.72X less than the CPU and GPU, respectively.

KW - FPGA

KW - Floating-point operation

KW - HPC

KW - OpenCL

UR - https://www.scopus.com/pages/publications/85042487448

U2 - 10.1007/978-3-319-75178-8_53

DO - 10.1007/978-3-319-75178-8_53

M3 - Conference contribution

AN - SCOPUS:85042487448

SN - 9783319751771

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

SP - 664

EP - 675

BT - Euro-Par 2017

A2 - Heras, Dora B.

A2 - Bouge, Luc

PB - Springer Verlag

Y2 - 28 August 2017 through 29 August 2017

ER -

Jin Z, Finkel H, Yoshii K, Cappello F. Evaluation of a floating-point intensive kernel on FPGA: A case study of geodesic distance kernel. In Heras DB, Bouge L, editors, Euro-Par 2017: Parallel Processing Workshops - Euro-Par 2017 International Workshops. Springer Verlag. 2018. p. 664-675. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-75178-8_53

Evaluation of a floating-point intensive kernel on FPGA: A case study of geodesic distance kernel

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this