Enhancing Memory Bandwidth in a Single Stream Computation with Multiple FPGAs

Antoniette Mondigo; Kentaro Sano; Hiroyuki Takizawa

doi:10.1109/FPT.2018.00078

Enhancing Memory Bandwidth in a Single Stream Computation with Multiple FPGAs

Antoniette Mondigo, Kentaro Sano, Hiroyuki Takizawa

Research and Development Divisions

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

1 Citation (Scopus)

Abstract

Stream computing is an area where FPGAs can be suitably utilized to meet high performance and high scalability demands. To achieve these, a deep computing pipeline is implemented on multiple FPGAs where stream computing is performed. This paper presents an approach to utilize two masters in a 1D ring network of multiple FPGAs for a single stream computation. Each master FPGA will be reading and writing to their respective DDR3 memories alternately, while streaming through the slave FPGAs. This is done in order to synchronize the computational results on physically separate memory units. Due to this, the aggregate memory bandwidth is doubled, which suggests enhanced performance. The introduction of this streaming concept lays the groundwork towards full utilization of memories in all the FPGAs, as an identified future work.

Original language	English
Title of host publication	Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	381-383
Number of pages	3
ISBN (Electronic)	9781728102139
DOIs	https://doi.org/10.1109/FPT.2018.00078
Publication status	Published - 2018 Dec
Event	17th International Conference on Field-Programmable Technology, FPT 2018 - Naha, Okinawa, Japan Duration: 2018 Dec 10 → 2018 Dec 14

Publication series

Name	Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018

Conference

Conference	17th International Conference on Field-Programmable Technology, FPT 2018
Country	Japan
City	Naha, Okinawa
Period	18/12/10 → 18/12/14

Keywords

1D ring topology
deep pipeline
enhanced memory bandwidth
high performance
multiple FPGAs
stream computing

ASJC Scopus subject areas

Software
Computer Science Applications
Hardware and Architecture

Access to Document

10.1109/FPT.2018.00078

Fingerprint Dive into the research topics of 'Enhancing Memory Bandwidth in a Single Stream Computation with Multiple FPGAs'. Together they form a unique fingerprint.

Cite this

Mondigo, A., Sano, K., & Takizawa, H. (2018). Enhancing Memory Bandwidth in a Single Stream Computation with Multiple FPGAs. In Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018 (pp. 381-383). [8742287] (Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/FPT.2018.00078

Enhancing Memory Bandwidth in a Single Stream Computation with Multiple FPGAs. / Mondigo, Antoniette; Sano, Kentaro; Takizawa, Hiroyuki.

Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018. Institute of Electrical and Electronics Engineers Inc., 2018. p. 381-383 8742287 (Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018).

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

Mondigo, A, Sano, K & Takizawa, H 2018, Enhancing Memory Bandwidth in a Single Stream Computation with Multiple FPGAs. in Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018., 8742287, Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018, Institute of Electrical and Electronics Engineers Inc., pp. 381-383, 17th International Conference on Field-Programmable Technology, FPT 2018, Naha, Okinawa, Japan, 18/12/10. https://doi.org/10.1109/FPT.2018.00078

Mondigo A, Sano K, Takizawa H. Enhancing Memory Bandwidth in a Single Stream Computation with Multiple FPGAs. In Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018. Institute of Electrical and Electronics Engineers Inc. 2018. p. 381-383. 8742287. (Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018). https://doi.org/10.1109/FPT.2018.00078

Mondigo, Antoniette ; Sano, Kentaro ; Takizawa, Hiroyuki. / Enhancing Memory Bandwidth in a Single Stream Computation with Multiple FPGAs. Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 381-383 (Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018).

@inproceedings{c9466a847fb1454ca6bf746eeda5133d,

title = "Enhancing Memory Bandwidth in a Single Stream Computation with Multiple FPGAs",

abstract = "Stream computing is an area where FPGAs can be suitably utilized to meet high performance and high scalability demands. To achieve these, a deep computing pipeline is implemented on multiple FPGAs where stream computing is performed. This paper presents an approach to utilize two masters in a 1D ring network of multiple FPGAs for a single stream computation. Each master FPGA will be reading and writing to their respective DDR3 memories alternately, while streaming through the slave FPGAs. This is done in order to synchronize the computational results on physically separate memory units. Due to this, the aggregate memory bandwidth is doubled, which suggests enhanced performance. The introduction of this streaming concept lays the groundwork towards full utilization of memories in all the FPGAs, as an identified future work.",

keywords = "1D ring topology, deep pipeline, enhanced memory bandwidth, high performance, multiple FPGAs, stream computing",

author = "Antoniette Mondigo and Kentaro Sano and Hiroyuki Takizawa",

year = "2018",

month = dec,

doi = "10.1109/FPT.2018.00078",

language = "English",

series = "Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018",

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

pages = "381--383",

booktitle = "Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018",

note = "17th International Conference on Field-Programmable Technology, FPT 2018 ; Conference date: 10-12-2018 Through 14-12-2018",

}

TY - GEN

T1 - Enhancing Memory Bandwidth in a Single Stream Computation with Multiple FPGAs

AU - Mondigo, Antoniette

AU - Sano, Kentaro

AU - Takizawa, Hiroyuki

PY - 2018/12

Y1 - 2018/12

N2 - Stream computing is an area where FPGAs can be suitably utilized to meet high performance and high scalability demands. To achieve these, a deep computing pipeline is implemented on multiple FPGAs where stream computing is performed. This paper presents an approach to utilize two masters in a 1D ring network of multiple FPGAs for a single stream computation. Each master FPGA will be reading and writing to their respective DDR3 memories alternately, while streaming through the slave FPGAs. This is done in order to synchronize the computational results on physically separate memory units. Due to this, the aggregate memory bandwidth is doubled, which suggests enhanced performance. The introduction of this streaming concept lays the groundwork towards full utilization of memories in all the FPGAs, as an identified future work.

AB - Stream computing is an area where FPGAs can be suitably utilized to meet high performance and high scalability demands. To achieve these, a deep computing pipeline is implemented on multiple FPGAs where stream computing is performed. This paper presents an approach to utilize two masters in a 1D ring network of multiple FPGAs for a single stream computation. Each master FPGA will be reading and writing to their respective DDR3 memories alternately, while streaming through the slave FPGAs. This is done in order to synchronize the computational results on physically separate memory units. Due to this, the aggregate memory bandwidth is doubled, which suggests enhanced performance. The introduction of this streaming concept lays the groundwork towards full utilization of memories in all the FPGAs, as an identified future work.

KW - 1D ring topology

KW - deep pipeline

KW - enhanced memory bandwidth

KW - high performance

KW - multiple FPGAs

KW - stream computing

UR - http://www.scopus.com/inward/record.url?scp=85068337204&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85068337204&partnerID=8YFLogxK

U2 - 10.1109/FPT.2018.00078

DO - 10.1109/FPT.2018.00078

M3 - Conference contribution

AN - SCOPUS:85068337204

T3 - Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018

SP - 381

EP - 383

BT - Proceedings - 2018 International Conference on Field-Programmable Technology, FPT 2018

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 17th International Conference on Field-Programmable Technology, FPT 2018

Y2 - 10 December 2018 through 14 December 2018

ER -