[SOLVED] Socket temp problems - AMD A10-6700

Feb 8, 2021
8
1
15
For about two years now, my CPU has been acting up and overheating whenever I use it for games such as GTA IV or more demanding ones.

My socket temperatures fluctuate wildly between 57 and over 80 degrees and after 15-30 minutes of demanding gameplay the PC reboots and asks me whether I want too boot into safe mode.

These past two years I've had my PC in the shop to get it checked out, had it cleaned, thermal paste replaced, swapped a motherboard and even replaced my cooler with a liquid-based one (ID Cooling Frostflow X 120 - I know, cheap, but best I could afford). Still no luck, the CPU still overheats.

Anyone knows whether anything else can be done? Is this CPU just doomed? Thanks!

CPU: AMD A10-6700
Motherboard: F2A85X-D3H
Case: Cooler Master HAF 912 Plus (1 x 200mm front fan, 2x 120mm top fans, 1 x 120mm side fan and the water cooling fan at the rear)
 
Solution
The motherboard I've recently replaced, bought used in November 2020. My previous motherboard was an ASUS F2A85-V PRO, had to be replaced for unrelated reasons, but the PC exhibited the same overheating symptoms with this one too, leading me to think it's not the mobo at fault.
__
@xravenxdota I'd love to hear if that fix works out for you. The problem sounds to be about the same as mine. Can't play any game without the PC restarting (no blue screens), but it doesn't seem hot to the touch and the core temps seem fine.
I am so sorry I haven't replied any sooner. The notification must have gotten past me. Thanks for bumping the thread by upvoting my last post.

The VRM should be fine with a...

Lutfij

Titan
Moderator
Welcome to the forums, newcomer!

You might want to see if reducing the in-game details and the resolution helps avert the high temps. You will also need to see if your motherboard is pending any BIOS updates. If you have a number of updates to go through, gradually work your way to the latest version.

Make and model of your PSU and it's age? In all honesty you should've gotten a 240mm AIO to cool that APU if you want to game with it, that or an airconditioned room.
 
Feb 8, 2021
8
1
15

Hello, thank you for the reply! My PSU is a Sirtec HPC-500-G12S (500W). I believe I've bought it in 2013 if memory serves, but I've had it tested in shop and I've been told it outputs alright. Maybe I should post a screenshot of my voltages here?

My room temp is ~20C, and even with all fans on the PC still reboots. The socket temp seems to be quite high from startup too (starts at ~50 on first boot).
 

MasterMadBones

Distinguished
It sounds to me like there are two separate issues. For the first one - temperature - I'd like for you to install AMD Overdrive, because that is the only software that can interpret the sensors on APUs from that era correctly. It reports a "thermal margin", which is how much hotter the CPU can get before it throttles or shuts down.

Do you get any kind of bluescreens or does it just randomly shut down?
 
Feb 8, 2021
8
1
15
Thank you for the reply. I've had Overdrive installed for quite a while. At the moment, my cores are sitting at 63-50 C thermal margin with nothing but Google Chrome open. When I'm playing games, it drops to about 30-20 C margin, but not more. Yet it still reboots.

No bluescreens I'm afraid. The video signal goes first for a moment (monitor says "no signal"), then the PC reboots and asks me whether I want to boot Windows in Safe Mode.
 

MasterMadBones

Distinguished
The Overdrive numbers look totally fine and the HWInfo64 temps look very familiar as well. Obviously the cores are not 16 degrees, because that would be below the ambient temperature. Socket temperature is what some motherboards use to control things like fan speed and VRM safety mechanisms, so I'd like to see a screenshot of HWInfo64 as well as the Overdrive Board Status tab under load.
 
Feb 8, 2021
8
1
15
The Overdrive numbers look totally fine and the HWInfo64 temps look very familiar as well. Obviously the cores are not 16 degrees, because that would be below the ambient temperature. Socket temperature is what some motherboards use to control things like fan speed and VRM safety mechanisms, so I'd like to see a screenshot of HWInfo64 as well as the Overdrive Board Status tab under load.

These are my temps after ~2 minutes of GTA IV. The temperatures spike and if I continue to play it's guaranteed to reboot.
 

MasterMadBones

Distinguished
This confirms what I thought. The CPU is obviously not at 80 degrees when the thermal margin is 40 degrees - it could never run safely at 110-120 degrees.

The socket temperature is separate from the package temperature and is reported my the motherboard. At 55 degrees peak it's totally fine and that should not affect how your motherboard runs the VRM.

This means that the system is shutting down due to another cause. Usually in situations like this it's the PSU. It may be able to output fine under constant load still, but large current spikes may cause it to trip the OCP or some of the voltage rails could drop below their minimums, which should prompt the motherboard to power down. Especially the latter is a problem that can arise with age and may not have been tested.

We'll need to know the remaining specs of your system to be able to determine the likelyhood of this being the case though.
 

xravenxdota

Reputable
Aug 26, 2017
435
66
4,990
Before we upgraded my brothers pc he had a A8-3870k i think.Can't remember all the parts me old.We have the same problem.For his pc it's not really the socket but the vrm's.I tried everything proper case cooling bios updates new thermal paste.The socket reads it's 80c but when you touch it it's cold.So once we upgrade his daughters pc i will take that pc get a heatsink most prob something like a PI heatsink.Next to the cpu where the vrm's should sit the pc are warm to the touch.

So all in all.I will see if the temp drop down after the installation.PC can't game because it restarts.This has been a problem for idk how long as you don't expect a new pc to run like this.We tried to RMA it and they said there's nothing wrong with the motherboard when new.Like wtf.I even recorded a video and showed it but according to them it still works...
 
Feb 8, 2021
8
1
15
My specs are:
CPU: AMD A10-6700
GPU: NVIDIA GeForce GTX 760 (2047 MB)
PSU: Sirtec HPC-500-G12S (500W) (bought 2013)
Motherboard: F2A85X-D3H
HDD: Western Digital WD10-EZEX (3 years old, no overheating, bad sectors or anything)
RAM: 12 GB DDR3, can't remember the brand
Case: Cooler Master HAF 912 Plus (1 x 200mm front fan, 2x 120mm top fans, 1 x 120mm side fan and the water cooling fan at the rear)

The motherboard I've recently replaced, bought used in November 2020. My previous motherboard was an ASUS F2A85-V PRO, had to be replaced for unrelated reasons, but the PC exhibited the same overheating symptoms with this one too, leading me to think it's not the mobo at fault.
__
@xravenxdota I'd love to hear if that fix works out for you. The problem sounds to be about the same as mine. Can't play any game without the PC restarting (no blue screens), but it doesn't seem hot to the touch and the core temps seem fine.
 

MasterMadBones

Distinguished
The motherboard I've recently replaced, bought used in November 2020. My previous motherboard was an ASUS F2A85-V PRO, had to be replaced for unrelated reasons, but the PC exhibited the same overheating symptoms with this one too, leading me to think it's not the mobo at fault.
__
@xravenxdota I'd love to hear if that fix works out for you. The problem sounds to be about the same as mine. Can't play any game without the PC restarting (no blue screens), but it doesn't seem hot to the touch and the core temps seem fine.
I am so sorry I haven't replied any sooner. The notification must have gotten past me. Thanks for bumping the thread by upvoting my last post.

The VRM should be fine with a 65W APU like yours, which is not to say it absolutely couldn't happen, but it would require a bad board design on both of the motherboards you've had.

12 GB of RAM is a rather unusual amount. Is it 8+4 or 4+4+4 and do all of the DIMMs have the same spec?

Another thing you could try is to run the PC wihout the GTX 760 for a while. Try to play a couple of games with the integrated graphics, even if the framerate is terrible. We just need to see if it still crashes under load without the extra power usage from the GPU.
 
Solution
Feb 8, 2021
8
1
15
I am so sorry I haven't replied any sooner. The notification must have gotten past me. Thanks for bumping the thread by upvoting my last post.

The VRM should be fine with a 65W APU like yours, which is not to say it absolutely couldn't happen, but it would require a bad board design on both of the motherboards you've had.

12 GB of RAM is a rather unusual amount. Is it 8+4 or 4+4+4 and do all of the DIMMs have the same spec?

Another thing you could try is to run the PC wihout the GTX 760 for a while. Try to play a couple of games with the integrated graphics, even if the framerate is terrible. We just need to see if it still crashes under load without the extra power usage from the GPU.

No worries, I appreciate your support.

I've got 2x 4GB Ram (Corsair XMS3 DDR3 1600MHz) and 2x 2GB RAM sticks, but I can't remember the brand at the moment (might be Kingmax). According to Speccy, the RAM setup is rated at 665 Mhz, so I assume the 2Gb ones are bottlenecking the Corsair ones.

I will try running some games without my GPU to see if that improves my situation and get back to the thread. Thank you!
 
  • Like
Reactions: MasterMadBones
Feb 8, 2021
8
1
15
@MasterMadBones Thank you for the heads up and sorry for not getting back to you yet. I've been studying for an exam and have been somewhat afraid to mess with my computer until the exam is done in case I might screw up something.

Would simply plugging the monitor into the APU make a difference to the PSU's power load or is removing the GPU required?
 

MasterMadBones

Distinguished
@MasterMadBones Thank you for the heads up and sorry for not getting back to you yet. I've been studying for an exam and have been somewhat afraid to mess with my computer until the exam is done in case I might screw up something.

Would simply plugging the monitor into the APU make a difference to the PSU's power load or is removing the GPU required?
No problem, good luck with your exam.

Just plugging the monitor into the APU is unlikely to help at the moment. Depending on the motherboard configuration, there is a chance you won't get any video output at all, or that the video output from the GPU is passed through to the APU. Even if it works as desired, you'd still need to uninstall the Nvidia driver and install the AMD one. AMD drivers don't always play well with swaps, so it's a good idea to wait until you have a little more time to solve any problems that may arise with that.