Parallel Algorithm Development

I am trying to figure out what is the best practice when developing a parallel algorithm for a data distribution model. What could be the pros and cons of block allocation versus cyclic distribution of data in memory. Any help would be greatly appreciated.

+3
source share
2 answers

Quinn's "Parallel Programming in C with MPI and OpenMP" offers many examples of different ways to distribute data in parallel programming. There's even a decision tree that helps you understand which approach is most convenient, depending on your needs.

+1

, , ( ) .

. - , (1 ) . , .

, .

+1

Source: https://habr.com/ru/post/1711412/


All Articles