Basic HTML version of Foils prepared April 22 2000

Foil 31 Sequential and Parallel Performance I

From Parallel FFT and use in PDE Solvers Computational Science Class CPS615 -- Winter Semester 2000. by Geoffrey C. Fox


Now to understand performance of general algorithms in parallel or in caches, one must understand storage of the arrays.
Note first that labeling is not necessarily same as storage
if p(l) is any permutation of l=0...d-1, then we can store
This can be used to improve sequential performance by ensuring that when a given digit l is being processed, its storage digit p(l) is in range 0? p(l) < L, where the cache can hold 2L entries



© 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 Mon Apr 24 2000