BIGADV + GPU client on same rig

[matsta31]

Hey,

I am curious if anyone is running BIGADV and a GPU client (or two) on the same rig ? I was running my gigabyte rig on all 8 cores with no GPU client at a 4.1 O/C without issue for 3 Bigadv WU. I was forced (!) to add my 9800 GX2 in the rig. I think I made a mistake, but I kept the BIGADV on 8 cores while running 2 GPU clients. I am now facing EUE on 1 of my GPU core and I lost a BIGADV WU to a BSOD yesterday night.

This morning before getting to work, I decided to ramp down the BIGADV to 7 cores and I did set the affinity on core 7 (I had a list form 0 to 7) for my single GPU client running this morning. (EUE Oon the other) (FAH@client.exe in my process list)

Should I do something else ? Should I question my O/C even if it was running stable for about 5 days on big adv when there was no GPU client on ?

[Arinoth]

When you gpu fold ontop of cpu it stresses your OC more then just cpu folding. If you BSOD'ed then it means your OC is not truely 100% stable and you'll need to bump up a few more little settings. Typically you should have no issue bigadv wu folding on all 8 cores and folding on 1+ gpus (as i do 1 cpu 2 gpu with priority set high for the gpus, normal for the cpu and no one is locked to a cpu core)

[phil3_66666]

my main run bigadv + 3 gpu client(all gpu client are stopped now because of heat in the room), but if you dont want to have a tpf of 36/37min download bill proccess 2 and when your not using the pc put vmware at "above normal"(i think is that in english...)

[matsta31]

Merci phil - Arinoth,

So you recommend when GPU + BIGADV folding at the same time, I should put the BIGADV at normal and the GPU at above norma, using the Bill 2 softwrel ? Will try this tonight, thnaks !

[Arinoth]

I've found after flushing out any little minor issues (SLI folding, 'unstable' OC) that having vmware at normal (running linuxrouter's image) and my gpus at high that it gives me the optimal ppd.

[stoanee]

Running bigWU+2 gpu with process lasso. BigWU set to low, gpu's set to normal. Seems to handle everything ok. (probably just jinxed myself...)

[Jon_di2]

Originally Posted by matsta31

Hey,
I am curious if anyone is running BIGADV and a GPU client (or two) on the same rig ?

I was reguarly running 2 gpu clients and 2 instances of VMware on my Q9450 (before the BigAVG's came out). I found that it froze up when running all four sometime between 1-4 days (it was really random). Never froze if I only did the VMwares or only the gpu's. I reduced my OC from 3.2 to 3.0 and it worked out great, however the ppd I got from having wmware and the gpu's running wasn't large enough to bother with the VMWare.

I'm running the BigAVG one right now without problems but I reduced the oc just in case a well. I'd suggest to just reduce the OC on your cpu (also vmware seems to just add to the stress, I'm not using it with the bigavg, but I really dont know much about the bigavg's).

[matsta31]

My oc was pretty aggressive ay 4.1 :)

since I did not have any problem whit only vmware, I tought I was stable. I suppose the GPU added stress it seems it is to much. I guess I'll scale back to 3.8 and see where it goes from there. When I find time, I'll try to go back to 4.1.. some OC fun !

[chrisk]

Running -smp 8 vmware with two GTX 285s. Using Priffinity to keep the GPUs higher than the vmware, but the vmware above normal. No issues at all so far. This rig also runs a tonne of other apps (as its my main rig. The folding rig awaits Fermiis)