A fast poisson solver for hybrid reconfigurable system

Authors:
Vitor Gomes;Haroldo Campos Velho;Andrea Charão
Affiliations:
Laboratory of Computing and Applied Mathematics, Brazilian Institute for Space Research, São José dos Campos, SP, Brazil;Laboratory of Computing and Applied Mathematics, Brazilian Institute for Space Research, São José dos Campos, SP, Brazil;Computer Systems Laboratory, Federal University of Santa Maria, Santa Maria, RS, Brazil
Venue:
ARC'13 Proceedings of the 9th international conference on Reconfigurable Computing: architectures, tools, and applications
Year:
2013

Citing 5
Cited 0

High-Performance Radix-2, 3 and 5 Parallel 1-D Complex FFT Algorithms for Distributed-Memory Parallel Computers

The Journal of Supercomputing
An Analysis of the Double-Precision Floating-Point FFT on FPGAs

FCCM '05 Proceedings of the 13th Annual IEEE Symposium on Field-Programmable Custom Computing Machines
High Performance Linear Algebra Operations on Reconfigurable Systems

SC '05 Proceedings of the 2005 ACM/IEEE conference on Supercomputing
Hardware/Software Integration for FPGA-based All-Pairs Shortest-Paths

FCCM '06 Proceedings of the 14th Annual IEEE Symposium on Field-Programmable Custom Computing Machines
State-of-the-art in heterogeneous computing

Scientific Programming

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents the design and implementation of a fast Poisson solver on a reconfigurable hybrid system. Our hybrid solver integrates a FPGA-based FFT coprocessor to collaborate in the solution of a numerical meteorological model involving one-dimensional shallow water equations. The Poisson equation is solved using a singular value decomposition associated with the Moore-Penrose inverse. The hybrid fast Poisson solver is evaluated under different amount of data entry and shows performance gains compared to the reference application.