Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network

  Advanced Search  

Landing CG on EARTH: A Case Study of FineGrained Multithreading on an Evolutionary Path

Summary: Landing CG on EARTH: A Case Study of Fine­Grained
Multithreading on an Evolutionary Path
Kevin B. Theobald, Gagan Agrawal, Rishi Kumar
Gerd Heber and Guang R. Gao
Department of Electrical and Computer Engineering /
Department of Computer and Information Sciences
University of Delaware
ftheobald, kumar,heber,ggaog@capsl.udel.edu; agrawal@cis.udel.edu
Paul Stodghill, Keshav Pingali
Department of Computer Science
Cornell University
Obtaining high performance from communication intensive irregular applications has been one of
the major hurdles in parallel scientific computing. This paper reports on our work in developing a
fine­grained multithreaded solution for the Conjugate Gradient (CG) problem.
There are two challenging and communication intensive steps in forming an efficient solution of
CG: performing the sparse matrix­vector multiply (MVM), and performing reductions followed by a
broadcast. In our recent work, we have developed a simple, yet very efficient, solution to executing
MVM on a multithreaded system. This paper present an efficient and effective mechanism for the


Source: Agrawal, Gagan - Department of Computer Science and Engineering, Ohio State University


Collections: Computer Technologies and Information Sciences