AMD FX 6100 Thermals Question

[pcformer]

Hello Everyone!

First of all, I'm new to the forum, and quite inexperienced in overclocking. I'm not educated on much of the terms used, however I do know a little.

That being beside the point, I have my BIOS set up with the clock at 3600MHz/3.6GHz and the voltage set to 1.35. Keep in mind I'm running the stock cooler. Now, this being said I also have Cool N Quiet disabled, hoping to keep the CPU fan at 100% all the time. After setting the clock in BIOS, when in windows, even at idle, the processor is running at 3600MHz.

My problem comes in when I put the computer under load. In my case, running GTA V on PC. When in game, the temp reading goes up to 180 degrees Celsius or more. This number seems insanely hot to me, however, even before forcing the clock at 3.6GHz in the BIOS, it was still reaching these temps when in game. Occasionally the processor reaches temps of 200+ and throttles down to 1.4GHz. I need a fix for this. Is the stock cooler really this bad? I would understand the temps if I was running a crazy overclock, however, I'm just running .3 more GHz, which with AMD Turbo **** enabled, it goes up to on it's own anyway. It just doesn't make very much sense to me because it seems as if it was reaching temps like this even on the stock clock under load? This just seems crazy to me. It's like the stock cooler is just completely useless at anything other than idle. The spec temp for the processor is 70.

What's the solution here? Just buy a better cooler? Is the chip bad? I'm lost.

Thanks in Advance Everyone!

 

[demien88]

Temp sensor is ****ed up. Your whole pc would burn down at 100c or shut down, idk exact numbers for fx6100 but fx6300 can go up to 65-70c. You need to tell people here exact configuration you use, everything plays role even case type fans attached to it, motherboard, cooler.
I asume you have everything at stock and some generic mobo. In that case I cannot recommend OC-ing

 

[Johan45]

Like demien said a list of your parts is quite useful and also a screenshot of something like HWMonitor free version while the PC is under load so we can see what's going on. It does sound like you're throttling but that 180° reading is false. Yes the stock coolers are not good for much over stock speed

 

[Blaylock]

to the Forums PCFormer!

As demien has mentioned it helps to see what other equipment you are running in this rig. If you can make a signature on your profile page similar to demiens it will only help us further.

Yes, one of two things is messed up. Either the sensor or the software reading the sensor. Most of us here use HWMonitor(free) for reading temps and voltages. Please download this and post a screen shot making sure to capture the voltages at the top along with the package and core CPU temps while under full load. Another very useful program is CPU-Z, also free, If you could show a screen shot of the CPU, memory, and SPD tabs in that program would give us more info.

The stock cooler is exactly that. It's designed to keep the CPU cool at stock clocks and voltages. If you wish to overclock it is highly recommended to get a quality aftermarket cooler. You don't need a top end one either. For a mild overclock a Cooler Master 212+ Evo or 212x is more than adequate. Just verify that it will fit in your case as they are rather tall. I prefer the 212x simply because it has a better fan and is also quieter.

The reason your CPU is remaining at 3.6GHz even at idle is because you disabled CnQ. CnQ allows the CPU to drop down at idle to keep it cool and the fans quiet. You actually want to disable this while fiqureing out your max overclock as it can cause some instability when pushing but will usually enable it again once we find our stable overclock speed.


EDIT: Ninja'd by Johan45 once again. LOL

 

[trents]

AMDOverdrive is AMD's in house temp/voltage monitoring app and it will give you a readout, not of the actual temp, but of the margin left before you hit the throttle down temp. This is what we call t he Tjmax which is the throttle down temp threshold. You might install this program as it should give the most accurate readings of the tools that are available.

Another thing we don't know is what temp reading your are referencing in your original post. "CPU" temp can be confusing as a temp label because in many or most software monitoring tools it refers not to the processor temp but the socket temp. More precise are temp readouts that are labeled "core temp" and "package" temp. These terms refer to what is actually happening with the thermals in the processor itself. Having said that, "CPU temp" or socket temp is also important as it reveals what is happening with the motherboard's power producing components which are also important. If these motherboard components are getting too hot they can cause the system to become unstable and shutdown.

What tool have you been using to monitor temps?

And as others have said, we really need to know more about your total system: GPU, PSU, Case, Memory, ventilation fans. You can see that all of us responding to you have our system components included automatically with our posts. That is called a Signature or "Sig" for short. Please create one for your system for us to refer to. You can use one of ours as an example of the information we are looking for. To create a Sig just click on "Settings" in the upper right corner of any forum page, scroll down until you find "Create Signature" on the left side column.

And yes, you have no hope of overclocking until you get a bigger cooler.

 

[pcformer]

AMDOverdrive is AMD's in house temp/voltage monitoring app and it will give you a readout, not of the actual temp, but of the margin left before you hit the throttle down temp. This is what we call t he Tjmax which is the throttle down temp threshold. You might install this program as it should give the most accurate readings of the tools that are available.

Another thing we don't know is what temp reading your are referencing in your original post. "CPU" temp can be confusing as a temp label because in many or most software monitoring tools it refers not to the processor temp but the socket temp. More precise are temp readouts that are labeled "core temp" and "package" temp. These terms refer to what is actually happening with the thermals in the processor itself. Having said that, "CPU temp" or socket temp is also important as it reveals what is happening with the motherboard's power producing components which are also important. If these motherboard components are getting too hot they can cause the system to become unstable and shutdown.

What tool have you been using to monitor temps?

And as others have said, we really need to know more about your total system: GPU, PSU, Case, Memory, ventilation fans. You can see that all of us responding to you have our system components included automatically with our posts. That is called a Signature or "Sig" for short. Please create one for your system for us to refer to. You can use one of ours as an example of the information we are looking for. To create a Sig just click on "Settings" in the upper right corner of any forum page, scroll down until you find "Create Signature" on the left side column.

And yes, you have no hope of overclocking until you get a bigger cooler.

Okay, so first things first, thanks for all the help and advice thus far. Keep in mind, I'm still very new to all of this. Used to working on cars lol. Been a while since I built the damn computer.

Okay, so after taking some of everyone's advice into consideration I've come to what I believe is a conclusion on a couple things. I believe the temp sensor on either the CPU or MOBO (whichever the temp reading comes from) is fault ONLY AFTER it exceeds 70°C. It seems that the temp readings are fairly normal and sensible at any temp before 70. For example, when idling the temp reading ranges between 35°C - 49°C. It seems though that once I get in game and start putting a load on the processor, allowing temps to increase, once the temp reading reaches 69°C-70°C, the reading SKYROCKETS to 200°C+. Then when the CPU throttles down to approximately 1.40GHz, the temps slowly start falling back down to the 60°C range for a bit. Then the process repeats itself, anytime the processor is under any fair load.

Also worth mentioning that I went out and bought this from Best Buy today. Replacing two of the pre installed case fans with these, and resting the third fan on the GPU facing the CPU, giving it a little more air.

Finally, I've screenshotted some temp readings and I think some of the stuff you guys asked for and attached them below. PLEASE let me know if you guys have any questions about them or need anything else. Really just wanting to solve this problem. Also worth mentioning I ordered the 212 EVO last night. Should be here either Saturday or Monday. I've also attached a list of my PC specs to my signature so you guys should be able to see those now.

Little Side Note: Any idea why my whole case, including everything in it (Hard Drives, GPU, Etc) starts to very lightly shock me if I touch them now? Seems as though something is shorting our or something making the whole case a big electric cube. Not sure if anyone could maybe have any ideas for this?

Pictures:::

 

[trents]

Sorry to sound so negative but the motherboard you are using is very inadequate for overclocking the FX-6100. It may be adequate for running the CPU at default frequencies and voltages and not give you any throttling but at default you don't have all the cores hitting turbo speed at the same time in order to keep heat and power draw down. You will need to invest not only in a better cooler but a better motherboard, one with at least a 6 phase power rating if you want to do more.

 

[pcformer]

Product suggestions?

 

[trents]

Absolutely! https://www.amazon.com/ASUS-970-Gam...1825282&sr=8-1&keywords=asus+am3++motherboard


or https://www.newegg.com/Product/Product.aspx?Item=N82E16813157479

The first board I referenced is a very good board that was reviewed by one of our forum experts and given high marks. Great value. Has 8+2 power phase I think.

As for the cooler: https://www.newegg.com/Product/Product.aspx?Item=9SIA4UF2DZ6565

How is it you have 10gb of Ram installed in motherboard with only two RAM slots? Do you have one 8 gb stick and one 2 gb stick?

 

[demien88]

Inspect your ram I suggest not to mismatch ram sticks because you can get into additional issues.
That "shock" you got is static electricity from components, it could lead to additional issues if its built up too much its most likely caused by case psu or motherboard.
Never had that issue so idk how to solve it maybe someone else here know how.
I suggest AURA mobo if you can afford it.

And yes sensors are screwed you can ignore those readings. There is no way cpu can reach 200c

 

[Blaylock]

I agree with Trents on the motherboard suggestion. The 212 will cool the gpu down considerably which could eliminate the throttling.

Something to keep an eye on but not related to your issues is your PSU is out of spec if the voltage sensors are to be believed. If you start experiencing random shut downs I would look into replacing it.

 

[trents]

Blalock, I fear his case will not give the clearance OP needs to install a Hyper 212 EVO which is 160mm tall. The Cryorig I suggested is about 154mm I think and cools at least as well as the CM 212 Evo.

 

[Blaylock]

Cryorg make some very good coolers too, no doubt. The specs on his case claims 160mm clearance so it's going to be a tight fit for sure. Hopefully it does because I believe he's already ordered it.

 

[trents]

Cryorg make some very good coolers too, no doubt. The specs on his case claims 160mm clearance so it's going to be a tight fit for sure. Hopefully it does because I believe he's already ordered it.

Did you get a PM from OP or something?

 

[Mr.Scott]

Also worth mentioning I ordered the 212 EVO last night. Should be here either Saturday or Monday. I've also attached a list of my PC specs to my signature so you guys should be able to see those now.

 

[caddi daddi]

I think I would like you to run the stability test in amd over drive and switch back to watch the temp in amd overdrive and see how close it gets to 0 degrees.
sometimes overdrives temp monitoring is better than others.

 

[trents]

I think I would like you to run the stability test in amd over drive and switch back to watch the temp in amd overdrive and see how close it gets to 0 degrees.
sometimes overdrives temp monitoring is better than others.

Ditto