PC freezes during gaming(HD 7770 overheating?)

Ray7600

Commendable
Oct 5, 2016
36
0
1,540
So I have been using an APU system(detailed specs at the bottom) for about 6 months and never had issues with it. Recently I borrowed my friend's old GPU, a Sapphire HD 7770 1GB D5 (this one http://www.sapphiretech.com/productdetial.asp?Pid=A14177C9-2709-4553-913F-F60364824EAF&lang=eng ) and have had freezing issues since then. The system would just randomly freeze, no keys nor mouse clicks worked and I had to manually power it down, and wait for a couple of minutes before it could boot back up. I tried troubleshooting myself and had some progress but the issue has not been completely resolved.

At first the system froze really quickly and randomly like 3-4 minutes into gaming, sometimes when I was simply browsing through my BIOS menu and once in the middle of re-installing drivers. Initially I thought it was because my Integrated Graphics had not been properly disabled yet and thus the iGPU and dGPU were conflicting. Sure enough, both GPUs showed up in the AMD Catalyst Control center and iGPU was set to 'Force' in BIOS. I disabled it, used DDU to do a clean uninstall and then re-installed drivers for the HD 7770.

The problem seemed partially solved as the system no longer froze until I was 10-15 minutes into heavy gaming. Performance seemed to noticeably improve as well(15-25 fps). So my next step was to clean the card the best I could(it was really dusty), wipe off the fan blades and the heat sink etc. and set up a Custom Fan curve through MSI Afterburner so as to run the fan at 100% when temps approached the 70s.

Then I ran Furmark GPU stress test and AMD Overdrive stability test simultaneously and observed that the GPU ran upto 89 degrees and after hovering in the high 80s for a minute it froze. Thus it seemed to confirm that overheating was the culprit.

However, yesterday while gaming for about 50 minutes, I was monitoring temps through Afterburner OSD. Temps did reach 84-85 but system was fine. And yet when it did freeze around the 50-minute mark, the GPU temp read only 73 degrees.

So now I am a bit confused as to what really is causing the system freeze. Any help to narrow down the cause or provide a solution would be highly appreciated.

I have tried clean re-installs of drivers but to no avail.

My system:
AMD A8-7600 Quad Core,
2×4GB Corsair 1600 Mhz DDR3 RAM,
Gigabyte F2A68HM-S1 motherboard,
Antec VP 550P PSU (bought only a week ago),
Toshiba 1TB HDD,
Sapphire HD 7770 1GB D5
Windows 10 64-bit

Games that froze: Assassin's Creed Revelations, Assassin's Creed Black Flag.
 
Solution
Hm, so we take PSU out the equation as it only froze and didn't crashed, try adding a fan aiming for the card or reapply thermal compound, maybe the card VRMs are overheating
Or, in the worst possible scenario, the core is damaged by extensive use at high temps (+80C) and that's why it keeps freezing

Elysian890

Commendable
Nov 9, 2016
189
0
1,760
I guess you used some good thermal compound on the GPU, 85C is high even for a 7770, PSU is fine for that build
If you have XMP enabled on your RAM, disable it and set the timings manually
Try to improve your airflow inside your case to prevent the card from overheating that much, you can add PCI slot fans to help remove the heat or remove the side panel, heck even blowing air with a household fan will help if there's no other way to keep it cool

Advice, try stressing with OCCT (free) using the Power Supply option, that will load your system to its max to see if it keeps failing, if it suddenly turns off while testing then that would be a PSU problem.

Nothing else comes to my mind, sorry if my answer isn't useful xx
 

Ray7600

Commendable
Oct 5, 2016
36
0
1,540
Thanks @Elysian890.
So you do think that the card overheating might be the issue?
Because last night it froze with the GPU at 73 degrees and that doesn't seem too high(it did go to the 85s earlier during the session but didn't freeze).
 

Ray7600

Commendable
Oct 5, 2016
36
0
1,540
@Elysian890 So I got OCCT and ran the power supply test as you recommended. As expected, after 20-25 minutes of full load, the system froze and I had to hard reset. Naturally I didn't get their graphs either. But this doesn't indicate a problem with the PSU I think... this was simply a repeat of what was already happening i.e. system freezing after being in full load for a while possibly due to overheating.
At the time of the freeze though, the GPU temp read 80 degrees.
 

Elysian890

Commendable
Nov 9, 2016
189
0
1,760
Hm, so we take PSU out the equation as it only froze and didn't crashed, try adding a fan aiming for the card or reapply thermal compound, maybe the card VRMs are overheating
Or, in the worst possible scenario, the core is damaged by extensive use at high temps (+80C) and that's why it keeps freezing
 
Solution