tang-hi commented on PR #12417: URL: https://github.com/apache/lucene/pull/12417#issuecomment-1629390714
> Some queries may benefit a lot from having the delta-decoding done in bulk for an entire block at once, and some may benefit from a lazy solution? Currently, my idea is to decompress directly without calculating prefix sums. Whenever the user needs it, we can compute a batch of values, but not all of them. For example, 16 ints (which align with the cache line). Perhaps through this approach, we can find a balance point. This is just my preliminary idea. Maybe I should submit another PR to test this idea? -- 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: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
