benwtrent opened a new issue, #13922:
URL: https://github.com/apache/lucene/issues/13922

   ### Description
   
   It would take a bit of refactoring, but:
   
   ```
       float dx = v - minQuantile;
       float dxc = Math.max(minQuantile, Math.min(maxQuantile, v)) - 
minQuantile;
       float dxs = scale * dxc;
       float dxq = Math.round(dxs) * alpha;
       if (dest != null) {
         dest[destIndex] = (byte) Math.round(dxs);
       }
       return minQuantile * (v - minQuantile / 2.0F) + (dx - dxq) * dxq;
   ```
   
   All these actions seem like they could be done lane-wise. Though, I expect 
the transformation from `FloatVector` to `byte[]` array to be the most 
expensive part and consequently, might make vectorizing this effectively 
impossible.
   
   of course, much of the stuff here written for value reuse would have to be 
adjusted


-- 
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.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

Reply via email to