GPU1 and GPU2 update

[Predrag Punosevac]

GPU1 is now rebooted. Memory should be cleared and zombie process 
killed. If something is not looking OK please email me. I am having 
troubles to figure out how to identify the slot of the GPU card which 
needs to be taken out of the server. I have the BUS locations and I know 
the BUS number of the failed GPU card. I just don't know if I can read 
off that BUS number from the card itself. I do not want to do trial and 
error method as pulling the card is very physically demanding. I will 
have to remove the server from the rack. I anticipate that I will have 
to remove GPU2 from the rack tomorrow morning.

PRedrag