linkedin facebook twitter rss

20 Mar granularity

This word is important in parallel computing and problem decomposition. When we solve a problem using a divide-and-conquer approach, granularity is the word we use to describe the coarseness or fineness of the pieces after we have divided the problem. The pieces of a computational problem computed on different processors of a parallel machine possess a granularity describing the coarseness or fineness of pieces of code that go to each processor. This is where modularity comes in: the finer the granularity of code modules, the closer we can approach a “knowledge driven” technical ecosystem.

Keywords: SIMD
MIMD
parallel computing
decomposition
References: algorithms
automata
computing
formalisms
logic
rules

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.