Presentation 2013-01-17
Implementation and performance evaluation of the accelerator for Lattice Boltzmann method on FPGA cluster
Yoshiaki KONO, Hayato SUZUKI, Ryotaro CHIBA, Kentaro SANO, Satoru YAMAMOTO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper reports prototype implementation and performance evaluation of a custom computing machine for the Lattice Boltzmann method (LBM) to be implemented on our tightly-coupled FPGA cluster. In general, strong scaling prevents performance from scaling with the number of computing nodes because it gets difficult to hide communication time with computing time. In our research, to solve this problem, we have proposed and developed the tightl-coupled FPGA cluster with an accelerator-domain network (ADN) to provide low-latency and high-bandwidth direct connection for FPGAs In this paper, we describe the architecute and stream computation of the LBM custom computing machine. We evaluate its performance based on prototype implementation with a single FPGA of the FPGA cluster.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) high-performance computation / lattice boltzmann method / accelerator-domain network / tightly-coupled FPGA cluster / FPGA
Paper # VLD2012-135,CPSY2012-84,RECONF2012-89
Date of Issue

Conference Information
Committee CPSY
Conference Date 2013/1/9(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Computer Systems (CPSY)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Implementation and performance evaluation of the accelerator for Lattice Boltzmann method on FPGA cluster
Sub Title (in English)
Keyword(1) high-performance computation
Keyword(2) lattice boltzmann method
Keyword(3) accelerator-domain network
Keyword(4) tightly-coupled FPGA cluster
Keyword(5) FPGA
1st Author's Name Yoshiaki KONO
1st Author's Affiliation Graduate School of Information Sciences, Tohoku University()
2nd Author's Name Hayato SUZUKI
2nd Author's Affiliation Graduate School of Information Sciences, Tohoku University
3rd Author's Name Ryotaro CHIBA
3rd Author's Affiliation Graduate School of Information Sciences, Tohoku University
4th Author's Name Kentaro SANO
4th Author's Affiliation Graduate School of Information Sciences, Tohoku University
5th Author's Name Satoru YAMAMOTO
5th Author's Affiliation Graduate School of Information Sciences, Tohoku University
Date 2013-01-17
Paper # VLD2012-135,CPSY2012-84,RECONF2012-89
Volume (vol) vol.112
Number (no) 376
Page pp.pp.-
#Pages 6
Date of Issue