icemelon9 edited a comment on issue #5214: [Runtime][Contrib] Support cudnn softmax URL: https://github.com/apache/incubator-tvm/pull/5214#issuecomment-608233532 @tqchen Yes, I understand that. But the latency difference could be 10x between tvm schedule and cudnn for the input shape like [100, 1024]. I guess to achieve such performance it requires fusion across multiple stage of reduction, which it seems not easy to be implemented in tir.
---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
