phaniarnab opened a new pull request #1471: URL: https://github.com/apache/systemds/pull/1471
This patch adds loop-tiling logic to the apply phase of transformencode to exploit CPU caches. Currently, the changes are limited to dense matrices. Loop-tiling shows 2x performance improvement in recoding a frame having 5M rows, 100 columens (100K unique in each) and w/ 32 threads. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
