Question CPU overheating

[skhoff27]

Hello, I have had a problem with my CPU for a few years: whenever the temperatures reach over 10°C (50°F) it overheats in certain cases:

1. Some non-demanding games.

2. Unzipping some zip or rar files.

3. Connecting to the internet and downloading Windows 10 updates.

4. Sometimes, staying idle for a few minutes, if I move the mouse the temperature stops rising and stabilizes again, reaching around 14°C (57°F).

5. When starting Windows.

6. When hibernating, in the last seconds before shutting down.

7. There are games in which it only overheats when a menu opens or there is a cinematic.

I've tried everything: keeping it clean, obviously, changing the thermal paste, adding fans to improve airflow, killing unneeded apps in background, I've never overclocked it, but nothing has worked, I have to wait until next fall to be able to play "heavy" games, which aren't that heavy either since my PC isn't very modern either, mid-2016.

Games like Wolfenstein The New Colossus, Ghost Recon Wildlands, Assassin's Creed Unity, Assassin's Creed Syndicate, Watch Dogs I and II, The Outer Worlds and Fallout 4 ran perfectly fine (in fall and winter seasons). I really don't get it, I'm really tired and bothered.

Specs:

CPU: AMD FX-8730 (Eight-Core) with Wraith Cooler

Motherboard: Gigabyte 970-Gaming

GPU: Nvidia GeForce GTX 750Ti

RAM: 12GB.

Thanks in advance, I would appreciate any help.

 

[Ralston18]

PSU: Make, model, wattage, age, condition (original to build, new, refurbished, used)?

Disk drive(s): make, model, capacity, how full?

Other attached devices.

 

[skhoff27]

PSU: Make, model, wattage, age, condition (original to build, new, refurbished, used)?

Disk drive(s): make, model, capacity, how full?

Other attached devices.

Disk drives: HDD: Western Digital WDC WD5000AZLX-00ZR6A0 500GB, two partitions, Drive E: 27GB used; Drive F: 61,5 GB used.
SSD: Kingston SUV400S37120G 111GB, 50GB used.
I can't see PSU make or model, but it was original to build, it's a Thermaltake.
The monitor is Samsung S20C300L.

 

[80251]

It sounds to me like the cooler should be dismounted from the CPU and a check of the TIM to see if it hasn't dried out or pumped out.
If your cooler is an AIO there's the distinct possibility the pump is dying.

 

[Phaaze88]

Those FX chips would report how far they were from throttling, not their actual operating temperature.
Also, is it really 10-14C inside your room..?

Motherboard: Gigabyte 970-Gaming

I suspect the VRMs on this motherboard have been taking a beating from that power hungry cpu; when the VRM power throttles, it throttles down the cpu with it. Some of the 970 chipset boards weren't rugged enough in power delivery to handle those 125w TDP-rated Bulldozer and Piledriver architecture cpus.

Perhaps dial down the core clocks or lower the power limit(or both). Should yield more performance in the scenarios the motherboard VRM is throttling the cpu. As for the scenarios where it isn't happening, yes, performance will be worse, but it is what it is.

 

[CountMike]

Hello, I have had a problem with my CPU for a few years: whenever the temperatures reach over 10°C (50°F) it overheats in certain cases:

1. Some non-demanding games.

2. Unzipping some zip or rar files.

3. Connecting to the internet and downloading Windows 10 updates.

4. Sometimes, staying idle for a few minutes, if I move the mouse the temperature stops rising and stabilizes again, reaching around 14°C (57°F).

5. When starting Windows.

6. When hibernating, in the last seconds before shutting down.

7. There are games in which it only overheats when a menu opens or there is a cinematic.

I've tried everything: keeping it clean, obviously, changing the thermal paste, adding fans to improve airflow, killing unneeded apps in background, I've never overclocked it, but nothing has worked, I have to wait until next fall to be able to play "heavy" games, which aren't that heavy either since my PC isn't very modern either, mid-2016.

Games like Wolfenstein The New Colossus, Ghost Recon Wildlands, Assassin's Creed Unity, Assassin's Creed Syndicate, Watch Dogs I and II, The Outer Worlds and Fallout 4 ran perfectly fine (in fall and winter seasons). I really don't get it, I'm really tired and bothered.

Specs:

CPU: AMD FX-8730 (Eight-Core) with Wraith Cooler

Motherboard: Gigabyte 970-Gaming

GPU: Nvidia GeForce GTX 750Ti

RAM: 12GB.

Thanks in advance, I would appreciate any help.

FX-8730 ??? Didn't you men 8370 ? Wraith cooler, unacceptable for 8core FX, How long has it been since paste was changed last time. What about case cooling ?
What are you testing and monitoring temperatures with ? Those temps you mentioned are practically impossible unless they are room temperatures‚ If for CPU, not even outside winter time in Alaska. Tjmax for that CPU is 60c core and 90c package temps.

 

[skhoff27]

It sounds to me like the cooler should be dismounted from the CPU and a check of the TIM to see if it hasn't dried out or pumped out.
If your cooler is an AIO there's the distinct possibility the pump is dying.

At most, the thermal paste must have been applied a year ago, maybe less. Anyway, as I said, it didn't work. It's not an AIO cooler.

 

[skhoff27]

Also, is it really 10-14C inside your room..?

Yes, that is the average temperature in autumn and winter. Why?

Perhaps dial down the core clocks or lower the power limit(or both).

How could I do it? I mean, I know there are apps like MSI Afterburner, but I don't know how much to reduce those parameters or which app is the most reliable.

I suspect the VRMs on this motherboard have been taking a beating from that power hungry cpu

So, it was working fine until a few years ago and now that it's almost ten years old, it's about to die? However, I don't understand why there were no problems with The Outer Worlds or Fallout 4 (both played this year) but there were with other games that are barely a couple of gigabytes in size and/or are 2D drawn with Paint, downloaded with the intention of having some distraction in the summer that doesn't kill my CPU.

 

[skhoff27]

FX-8730 ??? Didn't you men 8370 ?

Yes, I was wrong, it's 8370.

Wraith cooler, unacceptable for 8core FX,

It is the Cooler that came with the CPU.
https://www.watercoolinguk.co.uk/p/...th-Wraith-cooler-Socket-AM3--boxed_56729.html

How long has it been since paste was changed last time. What about case cooling ?

No more than a year, but possibly less. It has two fans to cool the case.

What are you testing and monitoring temperatures with ? Those temps you mentioned are practically impossible unless they are room temperatures‚ If for CPU, not even outside winter time in Alaska.

MSI Afterburner, Core Temp and Open Hardware Monitor.
Those are average room temps in fall and winter.

 

[CountMike]

Yes, I was wrong, it's 8370.

It is the Cooler that came with the CPU.
https://www.watercoolinguk.co.uk/p/...th-Wraith-cooler-Socket-AM3--boxed_56729.html

No more than a year, but possibly less. It has two fans to cool the case.

MSI Afterburner, Core Temp and Open Hardware Monitor.
Those are average room temps in fall and winter.

What are actual CPU temperatures ? Room temps don't say much. Is one or both over Tjmax I mentioned ? Those are just throttling temps, to shut down because of them they should be much higher.

 

[skhoff27]

What are actual CPU temperatures ? Room temps don't say much. Is one or both over Tjmax I mentioned ? Those are just throttling temps, to shut down because of them they should be much higher.

When idle, 14C, it is around 34C with peaks of up to 54C with some low-requirement games, sometimes there are peaks of up to 57C and when it exceeds them, it reaches 60C, 66C and more recently, 72C, the normal thing is that it descends quickly when the fan accelerates, but there are times when it does not stop rising and I have to close the application.
Core Temp states that the Tjmax is 80C, but on AMD's page it says the maximum is 61C.

 

[kanewolf]

When idle, 14C, it is around 34C with peaks of up to 54C with some low-requirement games, sometimes there are peaks of up to 57C and when it exceeds them, it reaches 60C, 66C and more recently, 72C, the normal thing is that it descends quickly when the fan accelerates, but there are times when it does not stop rising and I have to close the application.
Core Temp states that the Tjmax is 80C, but on AMD's page it says the maximum is 61C.

The temperature monitoring tools were bad about monitoring the AMD FX generation CPUs. AMD released a program called Overdrive It is so OLD that AMD no longer has it on their website. But you can find it at reasonably safe locations like -- https://www.guru3d.com/download/amd-overdrive-download
Overdrive doesn't show the CPU temp it shows the amount of temperature headroom left. So the display goes DOWN as the CPU heats up. When I had an FX CPU, that was the only software I used for monitoring.

 

[CountMike]

When idle, 14C, it is around 34C with peaks of up to 54C with some low-requirement games, sometimes there are peaks of up to 57C and when it exceeds them, it reaches 60C, 66C and more recently, 72C, the normal thing is that it descends quickly when the fan accelerates, but there are times when it does not stop rising and I have to close the application.
Core Temp states that the Tjmax is 80C, but on AMD's page it says the maximum is 61C.

That Wraith cooler couldn't keep decent temperatures on my FX 6350 it came with let alone my FX 8350 so I had to upgrade first to Cooler Master 212 evo and for 8350 to twin 120mmfan Mugen 3. So if undervolting doesn't help I suggest upgrading cooler.

 

[skhoff27]

The temperature monitoring tools were bad about monitoring the AMD FX generation CPUs. AMD released a program called Overdrive It is so OLD that AMD no longer has it on their website. But you can find it at reasonably safe locations like -- https://www.guru3d.com/download/amd-overdrive-download
Overdrive doesn't show the CPU temp it shows the amount of temperature headroom left. So the display goes DOWN as the CPU heats up. When I had an FX CPU, that was the only software I used for monitoring.

I will check this out.

 

[skhoff27]

So if undervolting doesn't help I suggest upgrading cooler.

What app do I need for undervolting? Or it's configured on motherboard setup?

 

[CountMike]

What app do I need for undervolting? Or it's configured on motherboard setup?

In BIOS lover max CPU voltage in 0.01v steps.

 

[skhoff27]

In BIOS lover max CPU voltage in 0.01v steps.

Well, I'll see if it works as soon as I can.

 

[skhoff27]

In BIOS lover max CPU voltage in 0.01v steps.

There is no option to set in 0.01v, the minimum seems to be +0.00v and the following number is 0.075v.

 

[Phaaze88]

Yes, that is the average temperature in autumn and winter. Why?

I'd be freezing if my room were at that temperatures like that. Perhaps you live and grew up in the colder climates, so you're accustomed to it. 20~23C is more the norm where I'm at for room temperature.

So, it was working fine until a few years ago and now that it's almost ten years old, it's about to die? However, I don't understand why there were no problems with The Outer Worlds or Fallout 4 (both played this year) but there were with other games that are barely a couple of gigabytes in size and/or are 2D drawn with Paint, downloaded with the intention of having some distraction in the summer that doesn't kill my CPU.

Electronics age, wear and tear like many other things. Hardware failures are inevitable, but if the parts paired together were already borderline ok, then things tend to just go south sooner... but shucks, skhoff27, your build is about 8 years old? That's pretty long.
Mine's about a year younger, and I feel like this has been long enough.

Games don't utilize your hardware's resources exactly the same, as they're simply not programmed that way.

Windows accumulates bloat over time, and it's good practice to clean install it from time to time(perhaps every other year), as it doesn't reliably wipe itself clean.