Gtx 980 ti shuts down entire computer but I don't know why

Hemi57L

Reputable
Nov 30, 2015
12
0
4,510
Computer is a lenovo k450 i5-4440 8gb ram and a brand new corsair cx650m. The system had the 980 ti before with a corsair cx600m and it ran fine, but now it would shut down the computer if there was high gpu load. I switched it with a gtx 960 for a few months and it had no issues. Thought it wasn't getting enough power so I got the 650 watt and put it back in but the issue still persists. The temps are fine, no artifacts, no lag, no overclock currently. Ran firestrike only if I had afterburner in the background with a custom fan profile (scored 18,203 for graphics) if afterburner isn't open it would shut down in the first scene. I tried resetting the bios, running the fans at higher speeds, and getting a different powersupply. It wont shut down while playing fortnite, but it will during cs:go and pubg. I'm not sure what else I can try. I do have a 980 ti in another system I can try switching them out another time. The card with the issue is an open box from microcenter, which means it was owned by someone else, could they have messed with the card's bios? It's an EVGA sc+. Don't remember if I had bought a protection plan.
 
Solution


Unless the issue is overheating, there is no way to fix a faulty video card unless you know how to check for bad components and replace them on the board, and it may be the actual processor that is the issue. Trying to flash the firmware won't hurt if the card is failing anyway. You can try testing it in...


No, it's the same 980 ti, it was working fine at one point. I'm starting to believe the voltage regulators are either busted or previous owner tweaked the card and we didn't notice and it damaged itself because we didn't have a fan curve on it. The card runs almost higher than 1.3v and it's boost clock goes up higher than 1303mhz but it's specified boost clock is 1190. I was so confused seeing the red bar in afterburner pass the red arrow, unless that's gpu boost 2.0 screwing with it. this card should only be around 1.1v. Card model is 06G-P4-4995-KR which is a sc+ running driver version 397.31 and the card's bios says Version84.0.41.0.90 exactly like that in the display adapters tab.

Update: Ran afterburner and Kombustor with the card in my computer. Fan speed was at 90% temps were at 75c voltage was at 1103mv but the card wouldn't hit boost and then the screen just went out and fans were at the same speed. Powersupply in this computer is a 750 watt evga g2
 


PSU is good quality, seems card is bad though. Try underclocking the card by 10% see if it is stable then.
 


It only goes down 90mhz and the issue still persists. Now I'm stuck because I can't find the receipt and I cut out the box for a rebate. If the card is bad an I have to do an rma what can I do? Not much. Should I just resort to a default vbios flash? The card doesn't have issues until there's high load, I'm using it right now, runs wallpaper engine just fine and I can browse the internet. Is there anything else I can try?

 


Unless the issue is overheating, there is no way to fix a faulty video card unless you know how to check for bad components and replace them on the board, and it may be the actual processor that is the issue. Trying to flash the firmware won't hurt if the card is failing anyway. You can try testing it in another system with a good PSU to make sure the issue follows the card.
 
Solution


Submitted a support ticket with EVGA, within a few minutes I got a reply telling me the card is faulty on a hardware level, just gonna have it RMA'ed. Thanks for letting me know sooner or I would've kept messing with it.