FPGA based custom computing machines for irregular problems

David Abramson, Paul Logothetis, Adam Postula, Marcus Randall

Research output: Contribution to conferencePaperResearchpeer-review

8 Citations (Scopus)

Abstract

Over the past few years there has been increased interest in building custom computing machines (CCMs) as a way of achieving very high performance on specific problems. The advent of high density field programmable gate arrays (FPGAs), in combination with new synthesis tools, have made it relatively easy to produce programmable custom machines without building specific hardware. In many cases, the performance achieved by a FPGA based custom computer is attributed to the exploitation of massive concurrency in the underlying application. In this paper we explore the sources of speedup for irregular problems in which is difficult to exploit such parallelism. We highlight 5 main sources of speedup that we have observed, namely the provision of high memory bandwidth, the use of flexible address generation hardware, the use of gather-scatter array operations, the use of lookup tables and the use of multiple tailored arithmetic units. By considering some representative examples of such irregular problems, the paper illustrates that good performance is possible given the current generation of FPGA devices and RISC processors. The paper then explores whether this performance gain will be possible given the next generation of RISC processors and FPGAs. It concludes that the only way to maintain the speedup is to alter the architecture of CCMs in combination with architectural changes to the FPGAs themselves.

Original languageEnglish
Pages324-333
Number of pages10
Publication statusPublished - 1 Jan 1998
Externally publishedYes
EventProceedings of the 1998 4th International Symposium on High-Performance Computer Architecture, HPCA - Las Vegas, NV, USA
Duration: 31 Jan 19984 Feb 1998

Conference

ConferenceProceedings of the 1998 4th International Symposium on High-Performance Computer Architecture, HPCA
CityLas Vegas, NV, USA
Period31/01/984/02/98

Fingerprint

Field programmable gate arrays (FPGA)
Reduced instruction set computing
Table lookup
Computer hardware
Hardware
Bandwidth
Data storage equipment

Cite this

Abramson, D., Logothetis, P., Postula, A., & Randall, M. (1998). FPGA based custom computing machines for irregular problems. 324-333. Paper presented at Proceedings of the 1998 4th International Symposium on High-Performance Computer Architecture, HPCA, Las Vegas, NV, USA, .
Abramson, David ; Logothetis, Paul ; Postula, Adam ; Randall, Marcus. / FPGA based custom computing machines for irregular problems. Paper presented at Proceedings of the 1998 4th International Symposium on High-Performance Computer Architecture, HPCA, Las Vegas, NV, USA, .10 p.
@conference{cf4f28651ee74b42b72acd88d1234da4,
title = "FPGA based custom computing machines for irregular problems",
abstract = "Over the past few years there has been increased interest in building custom computing machines (CCMs) as a way of achieving very high performance on specific problems. The advent of high density field programmable gate arrays (FPGAs), in combination with new synthesis tools, have made it relatively easy to produce programmable custom machines without building specific hardware. In many cases, the performance achieved by a FPGA based custom computer is attributed to the exploitation of massive concurrency in the underlying application. In this paper we explore the sources of speedup for irregular problems in which is difficult to exploit such parallelism. We highlight 5 main sources of speedup that we have observed, namely the provision of high memory bandwidth, the use of flexible address generation hardware, the use of gather-scatter array operations, the use of lookup tables and the use of multiple tailored arithmetic units. By considering some representative examples of such irregular problems, the paper illustrates that good performance is possible given the current generation of FPGA devices and RISC processors. The paper then explores whether this performance gain will be possible given the next generation of RISC processors and FPGAs. It concludes that the only way to maintain the speedup is to alter the architecture of CCMs in combination with architectural changes to the FPGAs themselves.",
author = "David Abramson and Paul Logothetis and Adam Postula and Marcus Randall",
year = "1998",
month = "1",
day = "1",
language = "English",
pages = "324--333",
note = "Proceedings of the 1998 4th International Symposium on High-Performance Computer Architecture, HPCA ; Conference date: 31-01-1998 Through 04-02-1998",

}

Abramson, D, Logothetis, P, Postula, A & Randall, M 1998, 'FPGA based custom computing machines for irregular problems' Paper presented at Proceedings of the 1998 4th International Symposium on High-Performance Computer Architecture, HPCA, Las Vegas, NV, USA, 31/01/98 - 4/02/98, pp. 324-333.

FPGA based custom computing machines for irregular problems. / Abramson, David; Logothetis, Paul; Postula, Adam; Randall, Marcus.

1998. 324-333 Paper presented at Proceedings of the 1998 4th International Symposium on High-Performance Computer Architecture, HPCA, Las Vegas, NV, USA, .

Research output: Contribution to conferencePaperResearchpeer-review

TY - CONF

T1 - FPGA based custom computing machines for irregular problems

AU - Abramson, David

AU - Logothetis, Paul

AU - Postula, Adam

AU - Randall, Marcus

PY - 1998/1/1

Y1 - 1998/1/1

N2 - Over the past few years there has been increased interest in building custom computing machines (CCMs) as a way of achieving very high performance on specific problems. The advent of high density field programmable gate arrays (FPGAs), in combination with new synthesis tools, have made it relatively easy to produce programmable custom machines without building specific hardware. In many cases, the performance achieved by a FPGA based custom computer is attributed to the exploitation of massive concurrency in the underlying application. In this paper we explore the sources of speedup for irregular problems in which is difficult to exploit such parallelism. We highlight 5 main sources of speedup that we have observed, namely the provision of high memory bandwidth, the use of flexible address generation hardware, the use of gather-scatter array operations, the use of lookup tables and the use of multiple tailored arithmetic units. By considering some representative examples of such irregular problems, the paper illustrates that good performance is possible given the current generation of FPGA devices and RISC processors. The paper then explores whether this performance gain will be possible given the next generation of RISC processors and FPGAs. It concludes that the only way to maintain the speedup is to alter the architecture of CCMs in combination with architectural changes to the FPGAs themselves.

AB - Over the past few years there has been increased interest in building custom computing machines (CCMs) as a way of achieving very high performance on specific problems. The advent of high density field programmable gate arrays (FPGAs), in combination with new synthesis tools, have made it relatively easy to produce programmable custom machines without building specific hardware. In many cases, the performance achieved by a FPGA based custom computer is attributed to the exploitation of massive concurrency in the underlying application. In this paper we explore the sources of speedup for irregular problems in which is difficult to exploit such parallelism. We highlight 5 main sources of speedup that we have observed, namely the provision of high memory bandwidth, the use of flexible address generation hardware, the use of gather-scatter array operations, the use of lookup tables and the use of multiple tailored arithmetic units. By considering some representative examples of such irregular problems, the paper illustrates that good performance is possible given the current generation of FPGA devices and RISC processors. The paper then explores whether this performance gain will be possible given the next generation of RISC processors and FPGAs. It concludes that the only way to maintain the speedup is to alter the architecture of CCMs in combination with architectural changes to the FPGAs themselves.

UR - http://www.scopus.com/inward/record.url?scp=0031605489&partnerID=8YFLogxK

M3 - Paper

SP - 324

EP - 333

ER -

Abramson D, Logothetis P, Postula A, Randall M. FPGA based custom computing machines for irregular problems. 1998. Paper presented at Proceedings of the 1998 4th International Symposium on High-Performance Computer Architecture, HPCA, Las Vegas, NV, USA, .