could my gpu be failing?

predimus81

Distinguished
Feb 26, 2013
41
0
18,530
i have 2 evga 1080 ftw 3 cards in sli, for about a month now and ive never overclocked them i am using a dell aw3418dw and a acer z35 for monitors. I was playing pubg and the monitor started to flicker on and off as if it was loosing the signal on my main monitor which was the dell and looking at the temps of the gpu i was concerned at a possible gpu starting to fail? below are the temps during the flickering

gpu 1 https://www.dropbox.com/s/pquldgfbrxltnlb/primary%20gpu.JPG?dl=0

gpu 2 https://www.dropbox.com/s/gijvagefzw4ji4c/second%20gpu.JPG?dl=0
 

predimus81

Distinguished
Feb 26, 2013
41
0
18,530

ill do that ive only had this monitor for 3 days just started acting up today.
 

predimus81

Distinguished
Feb 26, 2013
41
0
18,530


ive also noticed in game that there is some slight flickering in the display
 
Sep 2, 2018
6
0
20
I'd be more inclined to say you've a heat transfer problem rather than a failing GPU. While your main chip is below the limits for an NVIDIA card, your card memory and card power limits are pegged, which is most likely your flicker cause.

If the GPUs are air cooled, check that your fan on GPU 1 is able to spin (it looks like your fans are fine though based on the image so I'm leaning elsewhere), if good, remove the card and clean out the dust from the fins (a good strong blast of compressed air should do the trick) to get good heat transfer from metal to air. If that still fails under load after cleaning the fins then I'd be leaning towards failed thermal compound between the contact surfaces of the secondary chips and the heat sink.

If your warranty is still good, call for a return. If it's not under warranty, pull the fan, shroud, and heat sink and give the chip tops a good alcohol (90% or better isopropyl) scrub and apply some arctic silver and put her back together.

If it's liquid cooled, check for bubbles, check for flow, and good thermal contact between the block and the chips.

Your second GPU looks fine as far as temps and numbers go.
 

predimus81

Distinguished
Feb 26, 2013
41
0
18,530



played some pubg which was the game i was playing the second monitor doesn't turn on or off at all dont get any flickering while playing either

 

predimus81

Distinguished
Feb 26, 2013
41
0
18,530


checked the gpu's and there is no dust on them inside of the fans or on the card at all

 

Kashimi

Honorable
Apr 14, 2015
730
0
11,410
So you switched the primary and secondary monitor and the first primary monitor which had issues now wouldn't turn on, but the secondary monitor which is now the primary turns on and runs just fine?
 
Sep 2, 2018
6
0
20


Then based purely on your screen grabs, I'd personally be doing surgery on my card (ie, pulling the shroud, fan, and heatsink(s) and removing the old thermal paste and applying new).

I'm not overly concerned about your main chip temp, it's your card memory temp and card power temps that have me leaning that direction. NVIDIA's cutoff is 91C for the 10XX series cards (93C for previous series), anything higher and the card will drop out. The weak point for any card's thermal transfer medium are the power and memory chips...if the thermal contact is not placed exactly where the chip meets the heatsink, you get crappy thermal transfer which translates to issues like what you're seeing.

I've fried my fair-share of GPUs by fiddling with their thermal contacts trying for a better transfer.
 

predimus81

Distinguished
Feb 26, 2013
41
0
18,530


i unplugged the primary and soley runing the secondary played pubg for a hour for testing with no issues at all primary gpu running gup temp 76 c power temp 87 c memory temp but running at 2560 x 1080 to where the dell is a ips panel runs at 3440 x 1440p