980ti SLI problem

[AtmaDarkwolf]

3+ year old pc, only recently(last few months) issues started to rear up with random gpu fail(where video sometimes, but not always, recovers) during gaming (GTA online, etc)

 

For a while I was able to 'fix' the issue by underclocking gpu to 85%power limit, leading me to think that it was possibly the psu that was having issues.

 

More recently (just last week) it started happening regardless, but only in full screen gaming.

 

And today it happens all the time, and a few times when not even in any games.

 

I have run every possible benchmark and test I could and everything comes up 'fine' (IE the cards, pc, etc runs as would expected for its build) and ive also tested the ram with memtest (over a weekend, no errors)

 

I am at a total loss now and can only figure its the gpu's (both) failing, but I find that very suspicious since both are 'failing' identical (Ive tested both alone, swapped slots, etc, exact same issue when freezes / gpu crash each time) which originally lead me to think it was the PSU problem.

 

I have updated drivers (even tried rolling back to older ones) and it hasn't removed the problem.

 

Before I dump cash on new GPU's or PSU, can anyone give me any more possible ideas I could use to test and try to narrow down what is causing it?

 

 

(My option now i figure is just to rebuild pc and re-apply all thermal pastes, but am hoping to avoid that until I can track down what it is first)

 

Pc specs are i7 4790k 32gigs ram with dual 980ti's, I run windows 8.1 (rather not move to windows 10 if I can avoid it lol)

[Jurrunio]

ehm, not mentioning the PSU model?

 

check temperatures too, HWinfo's sensor mode and grab a lot of these readings.

[AtmaDarkwolf]

sorry, I ran HWinfo (And Msi afterburner) and never once saw temps go anywhere that would worry me, on anything. The most I've ever seen the GPU go to was 85deg(C) and that was only during the furmark test.

 

the below is my actual invoice list when I bought the pc/parts (I had them assemble it, which was new to me also since I always built mine myself before, but since the assembly was done for free i went ahead with it)

 

I also will attach a pic (taken with phone, excuse the glare) of the temps (on MSI afterburner, on 2nd monitor) when the last freeze hit earlier today.

1 x Asus MAXIMUS VII HERO w/ Dual DDR3 1600, 7.1 Audio, Gigabit LAN, CrossFireX / SLI 1 x Intel Core™ i7-4790K Processor, 4.00GHz w/ 8MB Cache   1 x Sandisk Ultra II Series Solid State Drive, 480GB (This has been replaced with new 850 EVO 500gb SSD, but not sure SSDs matter here) 1 x Cooler Master Nepton 280L Ultra High Performance Liquid CPU Cooler 1 x Cooler Master V Series V1200 Platinum 1200W Modular Power Supply 2 x EVGA D.GeForce GTX980 Ti SC+ ACX 1 x Corsair Vengeance LP 32GB DDR3 1600MHz CL10 Quad Channel Kit (4 x 8GB), Black

[Raudi_]

 Roll back drivers?

[AtmaDarkwolf]

tried that, even went back to a version from 2015 and same error.(i have downloads on my slave drive from old pc)

[AtmaDarkwolf]

well it sucks that it appears I'm the only person in all history of computers that has come up with this problem. thanks anyways.

[Mick Naughty]

I have it quite a bit. When ever my oc program would lower my voltage on one card. 

 

I have a custom bios and would run max power and voltage on them. 

 

If one was lower enough that the oc wasnt stage on that card, I’d get those kinda errors.