Problems with SMP VM and GPU2

J

jimbowdang

Limp Gawd
Joined
Jul 26, 2008
Messages
237
A few months ago I was running the nVidia GPU2 client alongside a single instance of Notfreds in VMWare Workstation modded to run on 4 cores. The setup worked on my system out of the box (ie. no goofy tricks/tweaks were needed to keep one client from starving the other). Back when those bad SMP WUs were going out I shut off the VM and ran WCG instead. However, lately I've been trying to get the VM going again but now whenever I run Notfreds the PPD on my GTX 275 go way down (like 3600 or less on a 353 unit). I don't know what could have changed on my system since before that would have caused this. Before my VM would hover around 89-92% and leave plenty CPU cycles for the GPU client but now my total CPU usage stays at 100%, with VMWare sucking up 98-99% of that. I've tried setting the priority on VMWare down and GPU2 up but no difference. Has something changed with the WUs that Stanford sends out? Right now Notfreds is crunching on: Project: 2671 (Run 56, Clone 97, Gen 105).
 
Make sure your GPU client is set to run at a slightly higher priority (through the F@H configuration options). Also, open up your folding.vmx files in Notepad and add these lines at the bottom:
Code: 
priority.grabbed = "idle"

priority.ungrabbed = "idle"
If that doesn't work, you should consider using a program like WinAFC or Bill2's Process Manager to automatically raise the priority of the Fahcore_11.exe and Fahcore_14.exe processes.
 
Thanks Zero, your first two suggestions didn't help but I was able to get it running smoothly again with WinAFC by following the instructions in the Notfreds guide. I guess it just seemed odd to me that I never needed it before.



 
WinAFC is the only way that I can consistenly get it to work across all platforms all the time.
 
You must log in or register to reply here.
Back
Top