• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

gpu folding on WHS 2011

T

Toconator

[H]ard DCOTM January 2026
Joined
Jul 8, 2005
Messages
878
Alright, lost a Q6600 boxen about a week ago so looking for ways to make up moar D without building another boxen. My Server mobo has 5 PCI-E slots so I thought I'd see if GPU folding was viable before throwing down any cash on a stack of new GPU's. Slapped a spare GT240 in it just to see the impact and all went fine - until the WU is done. Stalls after finish and won't d/l another. I can exit the client and restart and off to the races again until WU finishes. I'm still using the 6.41 systray client and it grabs a core 11 (obviously - non fermi) but always stalls. I selected prefer full performance in the driver and have no screen saver so there should be no power or interupt issues. Also assigned Run as Admin in the properties but it always stalls. Anyone else GPU folding on WHS 2011? ideas?
 
I folded in WHS 2011 with a GTX 660Ti and used the new client. Worked just fine.
 
Make sure fahcore and fahclient are unblocked in the firewall settings? I fold SMP and with a GPU on my WHS 2011 rig.
 
Dookey : well that sounds promising if I want to switch to new v7 client. Reason I'm still running v6 is that when I built the Server I copied over the current smp directory from the old one to save a juicy WU in progress. The original GPU I put in this box (GF7300) couldn't fold anyways. I'm also familiar with v6 and the others before it and been a little lazy getting into v7 which looks a little more convoluted to set up and configure.

Wyxe: l disabled Windows Firewal for private (work or home) network as it was interfering with some devices in the house a few months ago and I rely on my router anyways (VERY rarely surf from the Server). Besides if I quit and restart the client it has no problem getting another WU, it just stalls after sending the results. I'm folding smp on it no problem as well so that pretty much rules out a firewall issue.

I'm not losing a ton of points yet as the GT240 only gets 4.5K ppd at best, it's just a hassle having to constantly babysit it to keep it going. ppd might be small but it'll still turn in 12+ WU's per day if I can keep it running. Also is a test to see if it impacted the Server (multi-streaming HD and music can be occasionally heavy) and even if it would work well on the OS. This videocard folded fine on the old Q6600 rig on Win XP before the PSU blew up taking the mobo with it, and it test ran a game of League of Legends just fine on the workshop rig before I put it in so I can rule the card out I think. Puzzling ...
 
Update: switched over to v7 client. SMP ppd is actually just a titch higher but GPU folding won't run at all. Says CUDA not detected. gpu-z detects CUDA just fine however and restarting the old v6.41 clients works as well, just won't get another WU when finished. I'm still trying to figure out how to add v7 client to HFM so I can monitor from my main PC as well. Augh ! Think I'll hit the mojo and call it a nite. Like the FAH Control tho, looks alot nicer.
 
Toconator said:
Update: switched over to v7 client. SMP ppd is actually just a titch higher but GPU folding won't run at all. Says CUDA not detected. gpu-z detects CUDA just fine however and restarting the old v6.41 clients works as well, just won't get another WU when finished. I'm still trying to figure out how to add v7 client to HFM so I can monitor from my main PC as well. Augh ! Think I'll hit the mojo and call it a nite. Like the FAH Control tho, looks alot nicer.
Click to expand...

If I remember correctly I had to install the CUDA toolkit and then it worked.
 
Are you RDPing into the server computer at all? I know that can mess with gpu folding. Just a thought.
 
bigted said:
Are you RDPing into the server computer at all? I know that can mess with gpu folding. Just a thought.
Click to expand...

Usually when I have to perform regular tasks I RDP in, but for this setup and config of v7 and all the restarts of the gpu client I've had to hook up a spare wireless mouse and keyboard. GPU systray client won't start on any of my machines thru RDP, reports unsupported hardware or driver.

Funny thing is even tho it still won't run GPU client on v7, since I restarted the v6.41 client last night it has completed 5 in a row with no manual restarts. I updated the driver to 306.97 at this time as well so I wonder if there was an issue with the old one (it was quite old, 296.10 I think - I was running a GF7300 previously).

Also got HFM from my main PC to recognize the v7 client. As I elected to run the SMP as a service I had to restart the Server for the allow settings to take effect, simply restarting FAH Control as stated in the instructions was insufficient.

Bit of a pain to run 2 diff versions of FAH on the machine but at least it looks like it will work on WHS2011 which is the main thing. Once I get a new GPU I'll yank this GT240 and use it as a PhsX card somewhere and try to load this puppy up. Maybe a newer card will run on v7. Got an Enermax 720W in here, should be good for 3 decent mid-range GPU's along with a stack of storage drives ...
 
bigted said:
Are you RDPing into the server computer at all? I know that can mess with gpu folding. Just a thought.
Click to expand...

^^^ This. RDP breaks GPU folding. Use anything else you like for remote access, team viewer/logmein/VNC/other.

Also, are you trying to run GPU as a service? This will not work with any MS OS after XP due to security changes.
 
Yeah, I've used Teamviewer before. I'm not currently RDPing in. Installed another GPU, a client discarded GT420 and it won't run under v7.2.9 either. Both GPUs are reported as CUDA not detected in FAH Control. Both fold just fine with v6.41 (except now the GT240 is waiting for a WU, guess the Core 11's are in need of a refresh). v7.2.9 has grabbed a few WU's I've never seen before and are prob not avail to v6.34 SMP but so far the GPU folding aspect of it has been a bust with both a GT215 and a Fermi card. I was thinking maybe it had issues with older hardware and core 11's but apparently not. FYI: The GT420 gets 100-200 more ppd than the GT240 even though it has only half the shader count.
 
Update - multi-gpu folding on WHS 2011:

After adding a GT420 along with an installed GT240 the GT240 would not run. It would d/l a WU but not run or generate a unitinfo.txt . I uninstalled both cards and installed them one at a time, rebooting and reinstalling the drivers. (note: I am running v7.2.9 for smp on the cpu)

Starting with the GT420 in PCI-E slot 1. v7.2.9 saw it as gpu0 but reported CUDA not detected and didnt assign a folding slot to it. I config'd v6.41 with the -gpu 0 flag, verbosity 9 etc. and started the client = no problem. Installed the GT240 in PCI-E slot 3, v7 no Cuda etc. and config'd v6.41 with -gpu 1 etc. and started the client = won't run. FAH Control opens by default after every reboot so it was staring me in the face and I noticed that the GT420 was listed as gpu1 and the GT240 was gpu0 ? I swapped the GT240 into PCI-E slots 4 & 5 to no effect, stays gpu0. Don't know why WHS2011 arbitrarily switches the gpu's around but stopping the clients and reconfig them to match the gpu#'s has them both folding now. **Something for other WHS gpu folders to watch out for.**
 
You must log in or register to reply here.
Back
Top