viper7882 commented on issue #13040: Import and GPU init extremely slow on 
Windows
URL: 
https://github.com/apache/incubator-mxnet/issues/13040#issuecomment-438239629
 
 
   A simple program and memory profiling explains the reason of slowness.
   
   `import mxnet as mx
   
   if __name__ == '__main__':
       a = mx.nd.ones((2, 3), mx.gpu())
       b = a * 2 + 1
       print (b.asnumpy())`
   
   
![download](https://user-images.githubusercontent.com/28743821/48411418-759d3c80-e77c-11e8-9482-13e62fab690b.png)
   
   The peak memory usage has exceeded 2GB using this simple program. Likely 
someone has loaded huge amount of library contents into the RAM. My best guess 
would be cu92mkl as stated in [installation guide 
here](https://mxnet.incubator.apache.org/install/index.html?platform=Windows&language=Python&processor=GPU).
   
   Unless you always have vacant 2-4GB of RAM, else loading mxnet 1.3.0 will be 
slow due to Windows have to spend time to load from disk to RAM. Btw, I'm using 
SSD and still experience slowness.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on 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