• 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.

F@H Keeps crashing

K

kirk121

Limp Gawd
Joined
Sep 20, 2007
Messages
159
I ma having a constant crashing problem where f@h gpu keeps saying that my computer is unstable, I have tried changing the drivers for the video card with 3 different versions but I keep getting the same problems.

So I am starting to wonder if it could be that I have a ati motherboad and I am trying to run a nvidia gpu for folding. I have a asus M3A78 motherboard running a 9800gt video card. I am running the lates cuda drivers to see if that makes any difference.

The computer/video card is not overclocked it is just my file server/folding computer. As far as I know bios/chipset/windows/video card drivers, are all up todate. So I am at a loss as to what could keep causing F@H to crash on my video card.

And I am running windows vista 64bit

Update: I have been looking through the f@h log files and have seen alot of project: 5752's and 5765 are getting the unstable machine error.

Thanks
 
Last edited:
I found the same thing using a 9600gt i bought from best buy, I had to under clock the card a bit, actually back to stock speeds as Zotac felt the need to pre overclock it.
You clould underclock the card a bit to see if that helps you out a bit.
 
hmm.., in all honesty I have never had any problems using the F@H GPU2 client. I don't massively OC' my SP's (just about a 100 or so on both my 8800GS and my 9600GSO) The only problems I've ever had with the F@H GPU2 client was getting work, it seems irregardless of what of what I did. .I tried RAM tests, firewall tests, DSL tests. Internet hardware tests like switching cables. I even called my ISP to ask questions and complain and finally, out of unmitigated desperation, I sacrificed a few chickens to the computer Gods (only kiddin' about the chickens) :rolleyes: I don't want to hear the server down crap, I investigated that theory and it wasn't the problem. One minute it worked fine and the next it was like molasses getting a WU ( a couple of hours, I'd shut down and try at some unGodly hours) :eek:

Now on my 2x 24/7 DCing Q6600 G0 WC'er's I use the F@H GPU2 client and the WCG clients. (both very stable for 24/7 boxen) :p

It sounds to me like you might have some kind of hardware problem (RAM, temperature of your GPU, GPU card OC', PSU not up to snuff not suppling enough voltage to the card, bad or loose connection to the card, etc ;))

Please post whatever you find. I don't think it's the mobo 'cause everything else works (doesn't it :confused:)

DCing for the CURE
 
I dont think that the computer is unstable, I run one of the f@h processor core clients and dont have any problems. As for the temps I dont think I have seen them any higher than about 55c I know that is a little hot I am in the process of upgrading my water cooling so that I can have the video card water cooled as well. For now I am trying to down clock the core and memory 100MHz to see if that will make any diff.

I am also going to run some memory/ processor / video card stress tests to check and make sure that they system is stable.

Will post back if I find anything

thanks for the help so far:D
 
system unstable is refering to your gpu.. not the actual system its self.. its just a generic error code that they use..

if anything id just stress test the gpu.. shouldnt really have to mess with the cpu or ram..

and running a amd chipset with a nvidia card is not an issue.. i run an 8800GT on a 780G motherboard.. and a lot of people here use 790FX or GX motherboards with 4 PCI-e x16/x8 slots to run quad 9800GX2's and stuff..
 
sirmonkey1985 said:
system unstable is refering to your gpu.. not the actual system its self.. its just a generic error code that they use..

if anything id just stress test the gpu.. shouldnt really have to mess with the cpu or ram..

and running a amd chipset with a nvidia card is not an issue.. i run an 8800GT on a 780G motherboard.. and a lot of people here use 790FX or GX motherboards with 4 PCI-e x16/x8 slots to run quad 9800GX2's and stuff..
Click to expand...

Ya I just wanted to make sure that it is not any thing else, but I have gone as far as to run a second power supply for the video card and didnt work, tried different memory for the computer, down clocked the video card 200 MHz on both the memory and core, tried pointing a fan right at the video card but nothing seems to get it to complete even a single work unit.

Does any one know if there is a problem with the 9800gt being able to run folding?

I am really at a loss as to what could be the problem with F@H:confused: I have been running a cpu version as well and it is working just fine, if the video card can pass the benchmark and other stress tests I have no idea what to do next.:mad:

Also can anyone think of any patches that I could be missing that might affect the F@H gpu?
 
Last edited:
well no luck on anything I have tried, I have reinstalled the chipset drivers, video card drivers, F@H GPU program. So I am starting to think that it might be something with windows, because up till this last few weeks folding had been running just fine 24x7. And now I cant get it to run even one work unit.:mad::mad:
 
are re you running anything in the background that requires use of the gpu? when you reinstalled F@H gpu did you make sure to delete all the f@H folders? theres one in your application data folder and one in your program files folder.. also is it throwing up an error when it says its unstable? copy and paste your log file when it errors out and maybe we can pinpoint the issue..

9800GT shouldnt have any issues folding.. since its just a rebadged 8800GT..
 
sirmonkey1985 said:
are re you running anything in the background that requires use of the gpu? when you reinstalled F@H gpu did you make sure to delete all the f@H folders? theres one in your application data folder and one in your program files folder.. also is it throwing up an error when it says its unstable? copy and paste your log file when it errors out and maybe we can pinpoint the issue..

9800GT shouldnt have any issues folding.. since its just a rebadged 8800GT..
Click to expand...

I just reinstalled my os and am going to give it a day or so to see if it can complete any work units, if that does not work then I will post my log to see if there is any thing that can be done.

where it tells you error it says:
error: could not get length of results file work/wuresults_01.dat
error could not read unit 01 file. removing from queue.

And that seems to go for all failed work units, which in the last few weeks seem to be all of them.
 
on just about all of my GPU clients, I will get at least one EUE per day (sometimes more, but not enough to EUE limit me). There are, as others have said, a string of just plain bad WUs out there that won't complete even at stock :(
 
capreppy said:
on just about all of my GPU clients, I will get at least one EUE per day (sometimes more, but not enough to EUE limit me). There are, as others have said, a string of just plain bad WUs out there that won't complete even at stock :(
Click to expand...

But my problem is that I cant complete even one work unit, I keep hitting the 24 hour wait it fails to complete so many units. I can restart both the F@H gpu client and the computer and nothing will allow it to finish one. I dont think that in the last 2 weeks I have been able to finish anything.

I have even changed the os to server 2003 and it had worked on it before, and now at least at the stock overclocked speeds it still wont finish any.

So Right now I am running 3dmark 06 looped at the monitors max res and 8xAA 16xAF to stress the card some I am going to let it run through out the next day or so if it can complete it.:confused:
 
Ok so 3dmark ran for 12 hours all through the hot part of the day and had no problems, now it is starting to get a few work units that it is finally able to finish 2 or 3. So the problem must be on the work units being sent out or something on their end for it to suddenly start to work again.

Because like I said before it as been about 2 weeks since I was able to finish 1 work unit.

Update: I am still hitting the 24 stop I might finish 1-2 work units if I am lucky, most of the time I dont even get one I have to restart it mulitple times before I get one that will work
 
Last edited:
Well I have run the memtestg80 program and after 6000 loops it had 240 errors, so I am guessing it is the card. I guess that I will put in for an rma unless anyone has any other ideas.
 
Has any one had any problems since nvidia changed the voltage circuitry? By chance I happened to touch them near the additional power supply area on the back of the card and they seem to get very hot. Does anyone think it is possible that they are getting to hot and causing the card to become unstable while folding, and for it not to be such a big deal when running tests like 3dmark?
 
not sure if this was covered or not, but on my 3 way box, Ive had to lower my OC's to stock, and that didnt fix it, so they are back up to 1600, but EVERY time I get a 353, I get an EUE at the end of it, no other unit has a problem, temps are good enough, and the system has power, but the 353's run me flat almost every time

 
Vaulter98c said:
not sure if this was covered or not, but on my 3 way box, Ive had to lower my OC's to stock, and that didnt fix it, so they are back up to 1600, but EVERY time I get a 353, I get an EUE at the end of it, no other unit has a problem, temps are good enough, and the system has power, but the 353's run me flat almost every time

Click to expand...
I have tried going down 300 on both the core and memory, I just looked at the log and all failed they were the 5700's but the temps dont seem at least for the core to be getting to hot 40-55c range. So I dont know if it is the card or the cards temps or some other hardware or power or what any more:confused::mad::( I mean I have even tried running a dedicated 450w power supply to the card and it still fails, I have tried 2 os's vista and server 2003, tried different memory, I have looped 3dmark 06 for 12 hours with all setting maxed AA 8X and so on.
 
Last edited:
I just saw this thread so I may be a days late. My nvidia 8800 won't fold anything the last 3 days. Seems around the time I did an update and restart. I hit the EUE limit one day and now it attempts to get WU. I never did anything on the front end to overclock it, I just set up the GPU client and let it go.
 
it has to be faulty from the source, or a bad command between hardware>drivers>FAHcore>WU
 
I'm seeing a rare EUE and always on a 353 wu. (two different boxes)
I delete unitinfo.txt, queue.dat, and everything in the work folder. Even if it's another 353 that's downloaded, it will run and complete it.
 
Kowan said:
I'm seeing a rare EUE and always on a 353 wu. (two different boxes)
I delete unitinfo.txt, queue.dat, and everything in the work folder. Even if it's another 353 that's downloaded, it will run and complete it.
Click to expand...
FYI, all you need to delete is the queue.dat file to make the client download another workunit. Deleting the unitinfo.txt file and the work folder is unnecessary.
 
You must log in or register to reply here.
Back
Top