<div dir="ltr">Thank you for the advice, Srikanth. I'll try to start four X servers. And thank you Burlen for the scripts. I think my X11 configuration file will need to be changed.<br><br></div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, May 11, 2015 at 11:37 AM, Burlen Loring <span dir="ltr"><<a href="mailto:burlen.loring@gmail.com" target="_blank">burlen.loring@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div bgcolor="#FFFFFF" text="#000000">
btw, here is an <a href="https://github.com/burlen/pvserver-configs/blob/master/servers/nautilus/3.10.0-x/start_pvserver-x.qsub" target="_blank">batch
script</a> starting ParaView with pvservers assigned to GPU's
round robin, from the SGI UV 1000 with 8 GPU. The other tricky
aspect of running PV on multiple GPU is configuring X11. Here is the
<a href="https://github.com/burlen/pvserver-configs/blob/master/servers/nautilus/3.10.0-x/xorg.conf" target="_blank">xorg.conf</a>
on this system, each card has it's own device, screen, and layout. <br>
<br>
Hope it helps.<br>
Burlen<div><div class="h5"><br>
<br>
<div>On 05/10/2015 09:30 PM,
<a href="mailto:srikanth.nagella@stfc.ac.uk" target="_blank">srikanth.nagella@stfc.ac.uk</a> wrote:<br>
</div>
</div></div><blockquote type="cite">
<pre><div><div class="h5">Hi John,
You need to run 4 X servers, one X server per GPU card.
<a href="http://www.paraview.org/Wiki/Setting_up_a_ParaView_Server#Multiple_GPUs_Per_Nodehttp://www.paraview.org/Wiki/Setting_up_a_ParaView_Server%23Multiple_GPUs_Per_Node" target="_blank">http://www.paraview.org/Wiki/Setting_up_a_ParaView_Server#Multiple_GPUs_Per_Nodehttp://www.paraview.org/Wiki/Setting_up_a_ParaView_Server#Multiple_GPUs_Per_Node</a>
Sri
________________________________
From: ParaView [<a href="mailto:paraview-bounces@paraview.org" target="_blank">paraview-bounces@paraview.org</a>] on behalf of John Moore [<a href="mailto:johnpmooreiv@gmail.com" target="_blank">johnpmooreiv@gmail.com</a>]
Sent: 08 May 2015 17:03
To: <a href="mailto:paraview@paraview.org" target="_blank">paraview@paraview.org</a></div></div>
Subject: [Paraview] Multiple GPUs per node
I am trying to take advantage of all four GPUs on a 32 node Amazon ec2 cluster. However, I seem to only be able to utilize the first GPU.
I run pvserver as follows:
1) Manually start an X display with the command : sudo X :0 &
2) run pvserver with the following command (using OpenMPI's mpirun):
mpirun -np 16 pvserver -display :0.0 : -np 16 pvserver -display :0.1
When I try to connect the client, I get an error saying that the client could not connect to the remote display. Everything works fine using display :0.0 only.
Does anyone know what I might be doing wrong here? Thank you for your time!
John
_______________________________________________
Powered by <a href="http://www.kitware.com" target="_blank">www.kitware.com</a>
Visit other Kitware open-source projects at <a href="http://www.kitware.com/opensource/opensource.html" target="_blank">http://www.kitware.com/opensource/opensource.html</a>
Please keep messages on-topic and check the ParaView Wiki at: <a href="http://paraview.org/Wiki/ParaView" target="_blank">http://paraview.org/Wiki/ParaView</a>
Search the list archives at: <a href="http://markmail.org/search/?q=ParaView" target="_blank">http://markmail.org/search/?q=ParaView</a>
Follow this link to subscribe/unsubscribe:
<a href="http://public.kitware.com/mailman/listinfo/paraview" target="_blank">http://public.kitware.com/mailman/listinfo/paraview</a>
</pre>
</blockquote>
<br>
</div>
</blockquote></div><br></div>