In short, if your software does not exploit parallel processing techniques, then your code is limited to less than 2% of the potential performance of the processor. And this is just for a single processor - it is even more critical if the code has to run on a cluster or a supercomputer.
1 comment:
Interesting post. I have recently started a github repo for parallel algorithms. I initially implemented some sorting algorithms such as a Parallel Quicksort and a PSRS (Parallel Sorting by Random Samply) algorithm. I have seen some good results with the PSRS algorithm. They are designed to be a drop-in replacement for Java's built in sorting mechanism.
You can see the code here: https://github.com/broadbear/sort. At some point I will start on some graph algorithms (I've already implemented several serial versions).
Post a Comment