Premium
What computational scientists and engineers should know about parallelism and performance
Author(s) -
Pancake Cherri M.
Publication year - 1996
Publication title -
computer applications in engineering education
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.478
H-Index - 29
eISSN - 1099-0542
pISSN - 1061-3773
DOI - 10.1002/(sici)1099-0542(1996)4:2<145::aid-cae5>3.0.co;2-d
Subject(s) - parallelism (grammar) , computer science , task parallelism , data parallelism , parallel computing
Parallel computing sounds straightforward: Apply multiple CPUs to solve bigger, more complex problems and get results faster. Unfortunately, the experiences of computational scientists and engineers show that the price tag for parallelism is high. It is possible to spend months of effort, only to find that the parallel program runs slower on six CPUs than the original version did on one. Should you make the investment? This article provides practical rules of thumb for predicting if parallelism is likely to be worthwhile, given the nature of your application and the amount of effort you want to invest. © 1996 John Wiley & Sons, Inc.