FPGA based coprocessor to calculate the energy of dipolar system

Author

Danese, G. ; De Lotto, I. ; Leporati, F. ; Spelgatti, A.

Author_Institution

Dipt. di Inf. e Sistemistica, Pavia Univ., Italy

fYear

2002

fDate

2002

Firstpage

227

Lastpage

234

Abstract

In this paper, an FPGA-based coprocessor is presented, interfaced with the CPU to a computer through a PCI bus. It allows a faster evaluation of the energy of a crystal, which, computationally, is usually the heaviest part of a liquid-crystal Monte Carlo simulation of a dipolar system. The work is part of a global project aiming to design and build a parallel system made up of a cluster of accelerated workstations communicating via one of the most advanced modern network technologies. First estimations of performance show a big acceleration with respect to the execution of the same application on an Intel CPU-based motherboard

Keywords

Monte Carlo methods; coprocessors; dielectric polarisation; digital simulation; electric field effects; electric moments; field programmable gate arrays; liquid crystals; molecular dynamics method; peripheral interfaces; physics computing; software performance evaluation; special purpose computers; workstation clusters; CPU interface; FPGA-based coprocessor; Intel CPU-based motherboard; Monte Carlo simulation; PCI bus; accelerated workstation cluster; communication network; computational load; dipolar system energy calculation; liquid crystal energy; parallel system; performance estimation; Acceleration; Application software; Central Processing Unit; Computer networks; Coprocessors; Costs; Electrical capacitance tomography; Field programmable gate arrays; High performance computing; Workstations;

fLanguage

English

Publisher

ieee

Conference_Titel

Parallel, Distributed and Network-based Processing, 2002. Proceedings. 10th Euromicro Workshop on

Conference_Location

Canary Islands

Print_ISBN

0-7695-1444-8

Type

conf

DOI

10.1109/EMPDP.2002.994276

Filename

994276