CaptainDuke commented on a change in pull request #8479:
URL: https://github.com/apache/tvm/pull/8479#discussion_r673807110
##########
File path: python/tvm/topi/cuda/scatter.py
##########
@@ -787,42 +787,45 @@ def gen_ir(data_ptr, indices_ptr, updates_ptr, out_ptr):
for i in data_ptr.shape:
fused_shape *= i
- # For now we avoid parallizing over dimensions indexed by `indices` as
- # there may be repeated indices and hadling parallel accumulation can
- # be hard. So we parallelize over X_M .. X_{N-1} instead. This will
- # work well when these dimensions are large enough to saturate memory
- # bandwidth, but performance will be bad when these dimensions are
- # small.
Review comment:
Added
--
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]
