Intel Advisor (and Inspector) on

Intel Advisor (and Inspector) on

Hello everybody, this is my first post on this forum, I hope that this is the right question and this is a well-formed question.

I want to profile my application executed on a remote Xeon Phi with VTune, Advisor and Inspector to obtain the best performances. I could use the command line version

amplxe-cl
directly on the remote Xeon Phi machine (since Parallel Studio is installed there too), but I love the GUI and I would prefer to keep using it. 
I could run 

 

ssh -X $REMOTEADDRESS

and then call

amplxe-gui

but I don't understand why (if for a bad connection, or if the remote Xeon single core is weak) but executing the GUI version of VTune on the remote machine is so much slower than lunching it from my local machine.

Here it seems that you can your use local GUI version of VTune Amplifier on a remote target machine to profile it (via SSH). However, I can't find an analogous solution for Advisor or Inspector. However, what is the best solution for developing my parallel code locally, deploy and compile it on the remote machine and profile it on the remote machine from my local computer using these 3 profiler tools?

Thanks so much for your help!

1 post / 0 new
For more complete information about compiler optimizations, see our Optimization Notice.