Title :
Robust design methods for hardware accelerators for iterative algorithms in scientific computing
Author :
Kinsman, Adam B. ; Nicolici, Nicola
Author_Institution :
Dept. of Electr. & Comput. Eng., McMaster Univ., Hamilton, ON, Canada
Abstract :
The ubiquity of embedded systems of increasing complexity in domains like scientific computing requires computation on models whose complexity has grown beyond what is economical to manage purely in software to requiring hardware acceleration - a key part of which is selecting numerical data representations (bit-width allocation). To address the shortcomings of existing techniques when applied to scientific computing dataflows, we propose a methodology for determining custom hybrid fixed/floating-point data representations for iterative scientific computing applications.
Keywords :
data flow computing; data structures; embedded systems; iterative methods; mathematics computing; natural sciences computing; ubiquitous computing; bit-width allocation; embedded systems; hardware acceleration; hardware accelerators; hybrid fixed-floating-point data representations; iterative algorithms; iterative scientific computing applications; numerical data representations; robust design methods; scientific computing dataflows; Acceleration; Computational modeling; Design methodology; Embedded computing; Embedded software; Embedded system; Hardware; Iterative algorithms; Robustness; Scientific computing; Bit-width allocation; Satisfiability-Modulo Theory;
Conference_Titel :
Design Automation Conference (DAC), 2010 47th ACM/IEEE
Conference_Location :
Anaheim, CA
Print_ISBN :
978-1-4244-6677-1