Mystery Problem AMD APU 7860K

Saycho

Prominent
Jun 16, 2017
21
0
510
bought a used pc for someone and dear god how I regret it, but now it's done and I have to deal with it the specs :
Amd APU A10-7860k with Thermaltake Nic L32 cooler
asus a88Xm A
8 Gb 1666 Ram
Thermaletake lite power 450W
Amd Rx 550 4Gb Oc sapphire pulse
2 thermaltake case fans 120mm

When using the graphic card, games crash to desktop after a while, it could be 5min, or 3 hours+, even simple games, when I change something and want to check if it works I try wow burning crusade ( private server ) it runs smoothly 5 mins then it freezes, screen goes black and tabs to desktop, with fallout 4 I can sometimes play for hours, but then crashes to desktop with no error message.

When I go to windows event log I always find error 4101 amdkmdap

When using integrated GPU the games cause BSOD and report from whocrashed said something about spinning endlessly I don't remember well, I sometimes get freeze + blackscreen when browsing internet, I tried memtest, data lifeguard from WD for 3 hours and no errors were found, I tried the cpu stress test from AMD overdrive and for 1 hours and no error was found too, so there's the mystery! help me please!
 
Solution
That's a good thermal margin (even better than mine on load). So it can't be a thermal issue for the APU. We can discard thermal issues for the APU as a probable cause.

From my usage of WhoCrashed it should keep a log of what it discovers on your system. As long as they're not deleted it would at least bring up the details of the last crash. I believe it would help you to narrow down the issue you're experiencing. (I'm still of the mind it's usually RAM or drivers.)

PSU problems are hard to troubleshoot, as I understand it. As a preliminary check you could use a hardware monitor which has voltage information of the PSU. Something like HWMonitor will have this information. It's not a proper substitute for an actual electrical test, but...


What are the temps?
Did you reinstall windows?
The pc is not good and could be failing.
 

Saycho

Prominent
Jun 16, 2017
21
0
510
Hw monitor doesn't work well for me it shows 71° on idle which is almost max temp while many others show around 36-43° and amd overdrive shows 60° thermal margin
 
Thermal margin at load is more important than idle. From my own set up I anticipate it should be fine (perhaps 10~20 deg C thermal margin on a full load), but I would suggest double checking when the CPU is at full load, perhaps during the AMD Overdrive stress test.

You mentioned WhoCrashed, so it sounds like you should have something to work with with respect to BSOD; but you do not elaborate further. This may help to identify the issue. At the very least, you'd need the specific type of BSOD to narrow down to do this. (Most often it's a driver issue, but specifics will help identify it.)

The error code in Event Viewer... this has a broad explanation of it.

Personal experience suggests crash to desktops are usually something application specific. You could check the game forums which may have more detailed information specific to the game in question.

I would also double check which BIOS version you have.
For reference: https://www.asus.com/uk/Motherboards/A88XMA/HelpDesk_CPU/
It seems that particular APU needs the highest numbered BIOS (2902). At the very least it is something to eliminate as a probable cause of instability.
 

Saycho

Prominent
Jun 16, 2017
21
0
510
I remember when I did the stress test with amd overdrive the thermal margin was around 30-35°

I can't find the dump files relating the those BSOD since it was 10 days ago sorry :/

but if it's a PSU problem, wouldn't removing the gpu + case fans resolve it ? I did that and the problem persists

I have the latest version of the bios 3001 before the update the pc was very unstable
 
That's a good thermal margin (even better than mine on load). So it can't be a thermal issue for the APU. We can discard thermal issues for the APU as a probable cause.

From my usage of WhoCrashed it should keep a log of what it discovers on your system. As long as they're not deleted it would at least bring up the details of the last crash. I believe it would help you to narrow down the issue you're experiencing. (I'm still of the mind it's usually RAM or drivers.)

PSU problems are hard to troubleshoot, as I understand it. As a preliminary check you could use a hardware monitor which has voltage information of the PSU. Something like HWMonitor will have this information. It's not a proper substitute for an actual electrical test, but I like to think it could at least spot the most obvious deviations from what the actual voltage is supposed to be. If the PSU is faulty, it remains faulty even after removing some components (at least from my expectations).

BIOS update would eliminate a possible processor incompatibility, so we can discard that as a probable cause.

jaslion asked whether you reinstalled Windows. It may help with certain issues if they're software causes rather than hardware.
 
Solution

TRENDING THREADS