pengzhao-intel commented on issue #14173: [WIP] MXNet AMP (automatic mixed 
precision)
URL: https://github.com/apache/incubator-mxnet/pull/14173#issuecomment-476435228
 
 
   > We tested all (except for the newest pose estimation) of the networks from 
GluonCV and for all of them we saw between few to over 100% speedups running on 
V100, with typical speedup being over 50%, depending on what the relative cost 
of convolution/gemms are compared to the rest of the network.
   
   @ptrendx  TensorCore with FP16 provides the 8X speedup against FP32.  
   What is the gap between 2X in MXNet with 8X from TensorCore?

----------------------------------------------------------------
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

Reply via email to