wsl-inspur opened a new pull request #5485: URL: https://github.com/apache/incubator-tvm/pull/5485
- Optimization of Conv2d Winograd algorithm on Tensor Core for NHWC layout. - Winograd with tensor core outperforms original winograd algorithm for all the batchsizes. - However, performance of winograd is worse than conv2d for large batchsizes when Tensor Core were enabled for both. - Performance improvements of resnet50 are fairly good for small batchsizes. Please see RFC link below for details: [https://discuss.tvm.ai/t/rfc-tensor-core-optimization-of-winograd-conv2d-on-tensor-core/6543](url) @Hzfengsy @Laurawly @vinx13 @jwfromm Please help to review. ---------------------------------------------------------------- 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: us...@infra.apache.org
