Scripted HTML version of Foils prepared 29 December 1996

Foil 27 Performance Analysis of Matrix Multiplication

From CPS615-Linear Programming and Whirlwind Full Matrix Discussion Delivered Lectures of CPS615 Basic Simulation Track for Computational Science -- 5 Decemr 96. by Geoffrey C. Fox *
Secs 619.2
Time of matrix multiply
  • Time for pipe broadcast of A
  • Time to roll B
  • Time to calculate C
  • Total time
Efficiency is given by
Overhead, where :
  • If n = m2 is the grain size, then
  • with typical two dimensional character with
  • overhead proportional to in d dimensions

© Northeast Parallel Architectures Center, Syracuse University, npac@npac.syr.edu

If you have any comments about this server, send e-mail to webmaster@npac.syr.edu.

Page produced by wwwfoil on Thu Aug 14 1997