EVGA TITAN Black x3 Issues

[jedoc]

ok first post on here,

I have a custom built PC been having this issue , here are the specs first 

 

MSI Xpower AC  z97    Mother board 

Intel 4790k     processor 

32 GB Corsair Dominator Platinum 2400 mhz     Ram

Corsair HX 1200i     power source 

 

3  x  EVGA TITAN Black 6gb  GPU's     No SLI      EK waterblocks 

 

now heres the issue .... i am running 6 28inch 4k samsung screens ...( i'm a Day-trader not a gamer )

 

recently, while working ... my pc will for no apparent reason   freeze up and i get the " windows ran into an issue and needs to restart" (nvlddmkm.sys)  blue screen 

when it restarts the 2 screens connected to the one card dont work , but if i reinstall the nvida drivers it will work again 

 

sometimes it happens 3 times in a single day then it wont happen for weeks 

 

my question is , is there a way i can figure out what it causing the issue without ripping my system apart .... is it the mother board , is it the gpu or a another issue 

 

any help or ideas would me much appreciated , ive contacted nvidia a few times and they take so long to get back and forth

 

let me know your thoughts

 

 

Thanks in advance 

 

 

 

 

 

[AdamH96]

Not too sure in the subject but are there any overclocks on the cards or power limit increases on the cards? I run into Blue screens when I screw around with the power limit for some reason. Any artifacts on the screens?

[Electronics Wizardy]

Im gonna guess there is something wrong with that card. Try removing it and see if it works

 

 

Also if you don't need gpu power, id get something like a nvs 810. Its a much lower power card, single slot and still supports 8 4k monitors, and is cheaper.

[Mick Naughty]

So you know the issue is card related by posting here?
I assume you watch the temps, his res causes a lot of heat so maybe with them being old and needing new paste to have decent temps again. Just a guess as having 3 of them my self.

[jedoc]

10 minutes ago, Mick Naughty said:

So you know the issue is card related by posting here?
I assume you watch the temps, his res causes a lot of heat so maybe with them being old and needing new paste to have decent temps again. Just a guess as having 3 of them my self.

Yes, well I think it is, I was just looking for more ideas 

[jedoc]

12 minutes ago, Electronics Wizardy said:

Im gonna guess there is something wrong with that card. Try removing it and see if it works

 

 

Also if you don't need gpu power, id get something like a nvs 810. Its a much lower power card, single slot and still supports 8 4k monitors, and is cheaper.

Well that's exactly what I'm trying to get around doing lol... 

[jedoc]

15 minutes ago, AdamHolt96 said:

Not too sure in the subject but are there any overclocks on the cards or power limit increases on the cards? I run into Blue screens when I screw around with the power limit for some reason. Any artifacts on the screens?

No, no over clocks on the gpu's 

[jedoc]

Is it possible that a 1200 watt PSU is starving one of the cards maybe? 

[jedoc]

22 minutes ago, Mick Naughty said:

So you know the issue is card related by posting here?
I assume you watch the temps, his res causes a lot of heat so maybe with them being old and needing new paste to have decent temps again. Just a guess as having 3 of them my self.

What temps should I be running on these? 

[Mick Naughty]

Depends, what are they currently on with just sitting on the desktop?

With the high res bug bs, mine would idle at 60c bu that's with just the memory as the core would be at idle. Memory would run at max which would cause all the heat.

[jedoc]

42 minutes ago, Mick Naughty said:

Depends, what are they currently on with just sitting on the desktop?

With the high res bug bs, mine would idle at 60c bu that's with just the memory as the core would be at idle. Memory would run at max which would cause all the heat.

ok ... im running Kombuster stress tests right now ...so far highest ive seen is 61c

[Mick Naughty]

2 minutes ago, jedoc said:

ok ... im running Kombuster stress tests right now ...so far highest ive seen is 61c

So when are you having the issues?

[jedoc]

Just now, Mick Naughty said:

So when are you having the issues?

at random times 

sometime it wont happen for weeks ...which makes it super hard to diagnose  

[Mick Naughty]

Just now, jedoc said:

at random times 

sometime it wont happen for weeks ...which makes it super hard to diagnose  

Could be anything failing. Momery or software are the probable causes. I'd do a mem test one night. Then some sort of registry mechanic. 

 

I assume the drivers are all up today and the cards idle correctly. 

[jedoc]

18 minutes ago, Mick Naughty said:

Could be anything failing. Momery or software are the probable causes. I'd do a mem test one night. Then some sort of registry mechanic. 

 

I assume the drivers are all up today and the cards idle correctly. 

ok , ill do some other tests asap 

 

as far as the cards idling correctly  ...you tell me, here is a screenshot of everything at idle http://prntscr.com/eql7p2  the middle one it the card that throws the error 

[Mick Naughty]

3 minutes ago, jedoc said:

ok , ill do some other tests asap 

 

as far as the cards idling correctly  ...you tell me, here is a screenshot of everything at idle http://prntscr.com/eql7p2  the middle one it the card that throws the error 

Seems about right, temps are quite high for idle but a lot could effect that like all those monitors.

 

I just re read your op. Now that I think about it, I had the exact same issue. Mine crashed a bit more but it was the same thing popping up. I think it was still on windows 10 at the time of all the issues. I never use ddu but I did a clean install of the drivers. If you don't game, I would get an old one like December 1st of 2015. Took me about 2 weeks to get it to stop as I wasn't sure what was causing it. Seems like a certain windows update causes the system to freak out.

[jedoc]

ahh ok ...now i feel like im getting somewhere lol ..... the only thing i notice is that the card that fails always idles at .01 V less then the other 2 cards (check the pic) im also wondering if for some reason the one card is geting starved or something (maybe not likely just putting it out there for comment )  

[jedoc]

the driver thing makes sense tho .... ill have to give it a try ....thanks tho man you been a big help 

[Mick Naughty]

The voltage thing wouldn't be an issue. My cards do the sameness. They idle and boost at different volts. Just slightly off.  Sure they could use more power. 

 

I run precision xoc. I make sure I have power and voltage limit on the overlay so I can see what the cards are doing. Also running a modded bios.