GPU Crashing (Not Overheating)

Empichu

Reputable
Aug 10, 2014
10
0
4,510
GPU Crashes even tho i hardly ever go above 50 degrees on any game

My sound stays on for a couple more seconds all that dies to begin with is just the screen, i am able to hear sound and such for a couple more seconds before the system reboots
 
Solution
Sounds like it could be your PSU or a possible driver conflict. I would recommend that your first make sure that your BIOS and drivers are up to date.

What are your build details (CPU, GPU, PSU, mobo, DRAM)? What are your CPU temps when the crashes occur? Have you ever used a different GPU with the system (sometimes other graphics drivers or incompletely removed old drivers cause problems)?
Sounds like it could be your PSU or a possible driver conflict. I would recommend that your first make sure that your BIOS and drivers are up to date.

What are your build details (CPU, GPU, PSU, mobo, DRAM)? What are your CPU temps when the crashes occur? Have you ever used a different GPU with the system (sometimes other graphics drivers or incompletely removed old drivers cause problems)?
 
Solution

Empichu

Reputable
Aug 10, 2014
10
0
4,510


PSU: 750w power supply
GPU: GTX 770
CPU: I7-4770
and 16 gb of ram and as mentioned it never goes past 50 degrees not even leading up to a crash

I have already installed all the new drives for all my devices but it dosn't seem to have any relations to anything like that at all
 
Is this a new build or is this a new problem that's occurring on a build that was working fine previously? Does it only occur while playing games?

Are you getting a blue screen at all? Any error messages? Beeps or lights on your mobo?

Your first post implied that 50 degrees was your GPU temp. If that was for your CPU, what are your idle and load temps on your GPU?

Is your CPU and/or GPU overclocked?

What is the make and model of your PSU?

When you say you've already installed all the new drivers, do you mean from a support disk or that you've gone to the support webpages for your components and loaded the latest drivers?

What make and model of motherboard do you have? Which BIOS version is installed?

A remote diagnosis relies completely on the information that you post here - the better the info, the better the odds of getting a good diagnosis and solution. There are many reasons why you're GPU could be crashing, assuming that it is in fact your GPU causing the crash. Going down the list, some of the possible causes are driver conflicts, out-of-date BIOS, faulty mobo, faulty GPU, faulty PSU. The list goes on. As best as possible, you want to eliminate all the possible causes you can so you can concentrate on the probable causes.
 

Empichu

Reputable
Aug 10, 2014
10
0
4,510


No i dont get a blue screen it just cuts to black which i assume is do to the GPU crashing first since it's connected to the screen via a hdmi cable

I run at about 30-33 in idle temp on my gpu and as i said it'd max out at 50. CPU for the most part follows suit

my power supply is "Corsair CX 750W PSU"

First from support disks but since crashing started i've updated all from their websites

my motherboard is an MSI Z97 GAMING 5, Socket-1150

Most of the stuff in the PC is relativity new tho i had a hdd problem just a couple weeks back which i solved by reinstalling everything onto a new ssd

Thanks for helping me out by the way! :D

(Note i made a mistake previously i dont have a GTX 770 it's 780)
 
Do you have BIOS version 1.5 installed for your mobo?

In the interest of eliminating possibilities, even if the likelihood of the components being involved is small, I would download HWMonitor and check your PSU voltage outputs and verify that the CPU and GPU temps match whatever readings you're getting from mobo/mobo UI. You'll also want to check that your mobo is not overdoing your CPU core voltage, that you're DRAM speed is correct and that the DRAM voltage is correct. If you want you can post screenshots of the data.

I would also run MemTest86 to verify that you're not getting errors from your DRAM, and HDTune to check your drive status.

Do you have access to another PC that you can try your GPU on to see if it runs well on another system?

http://www.cpuid.com/softwares/hwmonitor.html
http://www.memtest86.com/download.htm
http://www.hdtune.com/download.html
 
Okay, most of those numbers look good for idle, but your +12V looks low. Try getting a screenshot with your system under load - your +12V should be in the 11.4V-12.6V. You can use something like Prime95 to stress the system, but I recommend AIDA64. Run the System Compatibility Test and give it just long enough to get your CPU up to max, then stop the test and screenshot your HWMonitor results and post.

http://www.aida64.com/downloads/aida64extreme460exe

 
It looks like your GPU is only guilty of causing crashes indirectly - it's trying to draw power to it's rated capacity, but your PSU can't provide it...Crash!

Is it under warranty? If so, call Corsair and request an RMA (Return of Merchandise Authorization). If it's out of warranty, you'll just have to buy a new one. Before you do this though, check to see if the voltage showing in BIOS matches what HWMonitor is. If you know you're way around a digital multimeter, you can check the actual voltage - this is the most accurate way to measure PSU power output.

DO NOT ATTEMPT TO FIX IT! There's enough juice inside the PSU to kill you.
 

Empichu

Reputable
Aug 10, 2014
10
0
4,510


Turns out my GPU was locked at a certain voltage level tried upping it through the GPU Tweaker(Software) that came with the card, no idea if it still crashes but i think it's sorted

Thanks for the help! :D

Edit

It might be worth adding that under stress i can get it to max out at 1200mv after the tweak (Just if anyone would be wondering) and that i'll write back on the thread if the issue is solved
 
Interesting. If it's not crashing with an increased voltage requirement, that points back to driver or setting conflict, i.e. if it's getting enough power at increased voltage requirement levels, it would tend to indicate that it was getting enough at the lower requirement level, too. I'll be interested to see if it somehow sorted itself out.
 
Apr 6, 2020
1
0
10
Recently had the same problem and it was definitely a power related issue. Replaced the PSU now and all fine. The PSU I had been using previously was actually new, but it only had one 6 pin PCIe power cable and the GPU needed two, so I had been using a 4 pin molex - PCIe adaptor. I strongly suspect that this was where the problem lay, but didn't fancy spending any more time playing around with it, so new PSU (with plenty of 6 pin PCIe cables) did the trick.