sun-dev commented on issue #14496: performance degradation from 1.3.1 to 1.4.0 URL: https://github.com/apache/incubator-mxnet/issues/14496#issuecomment-476279184 Just to clarify, for the data type change from 32bit to 64bit, we're talking about the tensor shape data type, not the actual tensor values. So for example, @apeforest's change allowed 64bit shape data type for a tensor with float elements. And this performance degradation for transpose was measured with pure-C++ (mshadow) implementations, not MKLDNN. @pengzhao-intel does MKLDNN support 64-bit tensor shapes? I cant imagine that MXNet's changes from 1.3.1 to 1.4.0 reduced the performance of MKLDNN operators, but has anyone verified if there was any issues for MKLDNN ops comapred to older versions?
---------------------------------------------------------------- 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 With regards, Apache Git Services
