Skip to content

Use LoadN for short BF16 dot product inputs instead of scalar conversion - #2703

Open
vekterli wants to merge 2 commits into
google:masterfrom
vekterli:vekterli/bfloat16-boundary-handling-in-prologue
Open

Use LoadN for short BF16 dot product inputs instead of scalar conversion#2703
vekterli wants to merge 2 commits into
google:masterfrom
vekterli:vekterli/bfloat16-boundary-handling-in-prologue

Only use `LoadN` fallback when compiling with GCC

df82cf0
Select commit
Loading
Failed to load commit list.
This check has been archived and is scheduled for deletion. Learn more about checks retention
Google CLA / cla/google succeeded Sep 26, 2025 in 1s

✅ All contributors are covered under a CLA with Google

See https://cla.developers.google.com/ for more info about Google's Contributor License Agreement (CLA).

ℹ️ Googlers: Go here to view more details and manage scans for this pull request.

Details

The following contributors were found for this pull request:

✅ df82cf0 Author: @vekterli <ve****li​@vespa.ai>

(Only the first commit for a unique contributor is listed.)

View more details on Google CLA