Hey everyone, 
 
I have dropped off quite a bit in folding. Been moving and settling into my first job. My computer has also just been a headache for the last 2 months. It just freezes randomly, idling or at load. 
 
I've had a thread going in the general hardware for a while trying to get some help. 
http://www.evga.com/forums/tm.aspx?m=1574003
 
if you guys could, review the thread i had going and maybe make some new suggestions. Please read through the whole thing as ive tried alot already and would prefer not to repeat myself like i had to in that thread because people just didnt read. 
 
I appreciate it fellas. Fold on!

Sorry to read about the problems you are having and it does get frustrating.  The first thing I thought about after reading about half-way through your thread was the OCZ SSD drive(s).  When I won that Origin PC (via the EVGA Scavenger Hunt) last year, it came with (2) OCZ 50GB Vertex 2 drives in Raid config.  After about 4 days of Folding on it, I was having lock-ups very similar to what you are seeing.  I contacted Origin and they said "yes, we have seen quite a bit of that in our systems and we have switched off the OCZ SSD drives".  They sent me a Western Digital Caviar Black, I installed Windows on it and the lockup problem was gone.  Before I installed that WD Disk, I had actually tried to just run each of the OCZ drives as a standalone SSD with Windows.  The flakiness followed with both SSDs all the way from lockups to not even being able to get through an installation of Win7.
 
I crap-canned the OCZ SSDs and haven't looked back.  Not saying all OCZ drives are bad, but when a system builder had to switch off them due to repeated issues with lockups, I took that as a signal.  Coupled with my own experience with (2) flaky drives, I decided not to put any more OCZ SSDs in my rigs.  Your unexplained problem with hangups across both sets of OCZ SSDs (original and replacement) may have simply been problems with two different OCZ SSDs.  That is what happened to me with my (2) SSDs.
 
Out of curiosity, were you ever able to upgrade the firmware on your OCZ SSDs?  I do recall reading some reports of that lock-up/hang-up issue (with certain OCZ drives) being corrected by a firmware update.

i havnt been able to upgrade the firmware. Although, i havnt tried using the linux or Dos tool that they have. Only the in windows tool, WHich always says firmware download failed. 
 
the kicker is, after i took the ssd out, my problem went away for about 2 weeks. Then it started again, using a 1 tb WD drive. So it is really strange that the problem stopped for a while and then started again with no change to the system once i got it working. 

If the problem went away for that long, then I'd examine what happened around/just before it re-emerged.  What changes (if any) were made to the hardware, drivers, software, or anything that could potentially affect the PC (including Windows updates).  Sometimes, with computers we face more than one problem that looks like a single problem because the symptom is the same.  Does the problem only surface when you are Folding?  If so, then it could be something in the system that can't take the stress of that application.  Hang in there, we have lots of good tech-minds here and I expect we'll get to the bottom of this thing.

No it doesnt happen just folding. It happens when the computer is just idling while im not using it, or under load playing games or folding. I havnt been able to discern any real rhyme or pattern to when it happens. I usually just come back to it after not using it for a little while, turn the monitor on to find it frozen.
 
Also, in the case of rebooting, sometimes it will freeze on windows startup or just crash and reboot itself. It seems as thogh there is no steady stability. But it will run fine for days at a time sometimes, and othertimes, it will freeze numerous time in one sitting where im trying to figure out the problem.
 
This is my 4th ish custom rig now and i have never had any issues with my previous 3.
 
It seems like most of these issues started when i added a second 580 to my rig for surround, but i havnt been able to pinpoint either 580 as the cause because i have been able to game and fold for hours on them, and just other times the rig freaks out.

Yeah, some of my weirdest lockups over the years, both while running, launching or shutting down Windows have been GPU/GPU driver related.  You could have actually been experiencing two issues at once, one with the SSD and another possible GPU-related problem where both can result in the same problem/result on the PC.  Maybe now it is time to refocus on the GPU/Drivers.  Have you tried running with just one GPU over a period of time to see what effect that has upon the problem?  Also, have you tried changing the GPU driver back to a release that may have more stability?  I don't have any experience with multiple monitors as in your setup, so maybe there is something to explore there that someone else can offer thoughts.

well, i havnt done any of the gpu testing individually pertaining to this freezing problem. I did do it shortly before this issue started trying to diagnose a no boot issue. Turns out that that issue was a melted 12v line on the 24 pin plug.
 
I will look into the driver deal. I have tried a few ersions now, but maybe i need to go back further. Just would like to stay relatively new to maintain performance inprovements and features in surround.
 
I still need to RMA my ssd and maybe that will help. some of the problem.
i appreciate you help Tex!

Makes me wonder if that power supply isn't quite right.  or it's the motherboard.
Motherboards that are sli and tri-sli certified should have to have an extra molex plug-in for the PCIE slots.  Not having one is just daft/cheap design.
 
If you have fried the 12v line on the 24 Pin, there could have been extra damage done to the motherboards 12v traces.
 
Or that power supply isn't up to running your rig.  850w is cutting it pretty fine for a rig of your spec.
 
Being on water makes trouble-shooting much more difficult. 
The first thing I would do is pull a video card and see if that fixes the problem.  If it does, it's more than likely your PSU can't handle the load.
 
 

i am no longer running the 850. i have upgraded to an AX1200 so the psu is a beast. 
 
the mobo was rma'd since the original issue and found to not be problematic. 

ok, that being the case
 
PSU = ok
M/B = ok
CPU = ok
GPU = to be tested (sucks, but you need to pull them out and test one at a time)
HDD - try different hard drive as primary boot
 
m/b drivers - any newer/older motherboard/chipset drivers you can try.
 
 
Other things that come to mind is a bad SATA socket on your motherboard and/or bad sata cables. 
 
 
 
 

ok, i finally have some time to work on this. I will be testing the gpus individually tomorrow. hopefully something positive happens. 
 
thanks for any help

Do you have a 4 pin molex on the board? Or one of these perhaps, http://www.evga.com/products/moreinfo.asp?pn=100-MB-PB01-BR , I needed one for my Classy for the added safety with SLI 580s. 
 
Other then that I'm lost. Seems like you've got a HDD to replace the SSD (I read somewhere it isn't good to fold on a SSD alone), most likely reformatted and installed a new copy of Windows to fix any registry errors, replaced the PSU, RMA'd the board. Hope you figure out something, or find something new to give us a better idea.

If your MB supports booting with the primary gpu not being in the first pcie slot, you can just pull the power cords from the gpus and test them one by one in their current slots. Saves ripping apart the water loop.
This should at least help finding if its one of your gpus.

I do have that powerboost item, i bought it after i fried my HX850. 
 
im not sure if my mobo supports booting with primary gpu in a slot other than the first slot. 
 
its fine though, i got some new fittings and needed to clean my loop anyways, so ill do it all at the same time. 

ok, so it is still early to make any conclusions, but i removed both 580's and plugged in a gts 450. Well there has been no freezing for a couple hours now, where as before, it would freeze pretty quick after starting to fold and such. 
 
so preliminary results point towards a faulty 580. i will continue testing with the 450, and if no freezing occurs, i will move on to testing each 580 one at a time to see which one causes the issue. 
 
it is a bummer to have to RMA a 580, pain in the butt.

sooo, after testing my gts 450, and then both 580's individually, i have not gotten my computer to freeze at all today, it seems that the issue lies with when the cards are both in the system, running. 
 
i am now confounded. 
 
could it really be a psu issue? i mean, the ax1200 shouldnt be stressed really at all...

test the 580's in SLI only....no 450 installed
 

An AX1200 should have no problem at all with your Rig.
 
Next step is re-installing the GPUs in SLI.
If the problem starts again, I'm going to point the
finger at a software problem.
 
You have verified that the GPUs work individually (I assume you tested them under load for extended periods - at least 12 hour each Folding should do the trick)
 
the only thing I can think of Hardware wise knowing the cards work individually, is a bad SLI contact on one of the cards.  have a close look at the contacts for damage and give them a thorough clean.
 

Ok, well i put both cards back in. They have been running for about 5 hours now and no freezing has happened. 
 
Im a bit baffled a this point. I had reseated the cards before that didnt make a difference. 
 
I havnt put on the SLI bridge yet, not sure if that makes a difference. 
 
recently, the freezing happens pretty soon after turning the computer on. it never made it past an hour without freezing, usually only a few minutes. 
 
i don't like it when things just start working for no apparent reason, im worried its a fluke or something and will really irritate me if it starts back up. 
 
the good thing is, im finally getting some points racked up for this month. 

ok, figures, right after i post about it not freezing, it goes and freezes. 
 
so, after about 5 hours, it froze while folding on cpu and both 580's
 
so it seems that having both cards in the system caused freezing but not one at a time. what does this mean? 

no new ideas here?
 
Ive been running with one 580 unplugged since monday morning without so much as a hiccup. 24/7 folding going on with just one 580 running.

I'm just thinkin' out loud and don't know what the solution is (yet), but we're still here and want to help ya get those 580s Foldin'. 
 
I have a few thoughts. 

• Are the GPU's EVGA brand by chance?  If so, I'd lodge a call over to their Tech Support line to see what suggestions they may have for the problem.
• I don't exactly recall what equipment you are running and it may be of help to others who read this thread to briefly outline those things (mobo, PSU, Overclocks if in use, NVidia driver release, what else are you doing while Folding that could have impact, etc).  Someone may see that info and it may spawn an idea based on experience they may have with similar products.
• Since the problem seems to manifest itself only when both 580 GPUs are Folding coupled with CPU Folding, I'm wondering if there is some overall stress point being reached with processing.  I'm wondering if holding back on some of the CPU cores to allow the GPU's to have more bandwidth would yield any positive results.  You know, that Folding argument you can add that says "-smp xx" where xx is the number of CPU threads/cores that you will allow for SMP Folding.  Just a thought...

 

Thanks for the help!
 
I have updated my SIG with rig components. 
 
They are EVGA cards. so ill try giving Tech support a call. 
 
I am only running the folding client while trying to get this issue fixed. 
Ill try freeing up a core for the cards to use. 

I've had good experiences calling EVGA tech support for advice/help in the past, so I hope they may be able to shed some light.  Let us know how things go because we have a tremendous knowledge-base here within the Folding community and hopefully others will have some more ideas if needed.

cblaxx


ok, figures, right after i post about it not freezing, it goes and freezes. 

so, after about 5 hours, it froze while folding on cpu and both 580's

so it seems that having both cards in the system caused freezing but not one at a time. what does this mean? 

 
Did you have the SLI bridge installed when this failure happened?
 
 

Simba123

cblaxx


ok, figures, right after i post about it not freezing, it goes and freezes. 

so, after about 5 hours, it froze while folding on cpu and both 580's

so it seems that having both cards in the system caused freezing but not one at a time. what does this mean? 

Did you have the SLI bridge installed when this failure happened?

 
No, i didnt, I didnt have SLI enabled in the control panel either, so the cards should have been completely independent. 

Honestly I can only gander and say the PSU is the fault, or the video card drivers were not installed correctly at this point. If your saying the system freezes only when both cards are in the rig, and not 1 at a time, that sounds like the PSU to me. 

so is there any way to test the psu without sending it in for RMA? i dont want to pay to ship a beastly psu if it isnt for sure the issue...

I think there is I just forget the name of the tool you need >___> voltmeter? I may be wrong so hope someone comes in here to shine some light.
 
Also, have you tried the 580s in different slots? It may be one of the boards pci slots having an issue or something. 

i havnt, but if i use another slot, it becomes an 8x slot and in the past, i have had issues using 8x slots on boards. i havnt tried it on this board though. 
 
and i dont have an sli bridge(for my water loop) that will span more than one slot 
 
and i figured i could use a voltmeter, but i am not sure what to look for since the psu works perfectly for fully loaded cpu and OC'd 580.