How hot should the MOBO DRM's Be?

[Solarstarshines]

Is it normal for the chipset and Vrm heatsinks beeing hot to the touch?

Temps should be in the 20 to low 30C when lightly loaded. With a heavy load I would like to be no higher then 50-55C. GPU and GPU VRM's can deal with higher temps. Your 8350 should never be allowed higher than 61C. I don't know about the 970 chips but the 990 chip set can have high NB temps, lot of users add a tiny heatsink/fan for added cooling. Some add a 50-80mm fan to the back side of the board to cool cpu and vrm's.
View attachment 311469

Why shouldnt i let the 8350 go over 61c?
Also without overclock the Noctua Nh u14s keeps it at no more than 45c at 99% load

Why because the chip will degrade going higher The threshold is only that much due to the TDP being 125W
If it was 95 w or 77 like intel you could take it to higher temps theory here is less TDP Thermal Design Power cooler overclocks Thermal design power - Wikipedia, the free encyclopedia

 

[NoN]

I can feel you insecure and worried about those results.....You've got a very good CPU but an average board that can permit you pushing a little but not enough comparing to your CPU Specs.

I would only recommend you to increase the Airflow in your PC Case before being too worried about Temps.

 

[sisrace]

Yes could be...!

Also... I have been checking out some stuff on my mobo..
Does the Hypertransport have anything to do with overclocking? becouse this board has 4800MT/s
Does it mean i can overclock my 8350 to 4.8Ghz?
Becouse if then thats REALLY good considering this is a 970 chipset that isnt made for overclocking at all but it does support it. So i used to run my cpu at 4.2 ghz but i removed it when my psu whent bad and i havent set it back to the 4.2Ghz even now when i got a RM650W psu

First of all why did you OC it to 4.2 it does that on stock turbo

In a sense you didn't overclock anything you just maxed out the stock configuration any more and you would of ran into trouble most boards for that chip are 8+2 phase out of the box it's 4 ghz turbo 4.2

Saw that it turboed up to 4.2 so yeah. but it went up to 4.3-4.4 when stressing it while it was oc'd so i dont know what its doing

 

[sisrace]

Temps should be in the 20 to low 30C when lightly loaded. With a heavy load I would like to be no higher then 50-55C. GPU and GPU VRM's can deal with higher temps. Your 8350 should never be allowed higher than 61C. I don't know about the 970 chips but the 990 chip set can have high NB temps, lot of users add a tiny heatsink/fan for added cooling. Some add a 50-80mm fan to the back side of the board to cool cpu and vrm's.
View attachment 311469

Why shouldnt i let the 8350 go over 61c?
Also without overclock the Noctua Nh u14s keeps it at no more than 45c at 99% load

Why because the chip will degrade going higher The threshold is only that much due to the TDP being 125W
If it was 95 w or 77 like intel you could take it to higher temps theory here is less TDP Thermal Design Power cooler overclocks Thermal design power - Wikipedia, the free encyclopedia

but but.. The thermal threshold of the 8350 is 80c ?

 

[Solarstarshines]

Why shouldnt i let the 8350 go over 61c?
Also without overclock the Noctua Nh u14s keeps it at no more than 45c at 99% load

Why because the chip will degrade going higher The threshold is only that much due to the TDP being 125W
If it was 95 w or 77 like intel you could take it to higher temps theory here is less TDP Thermal Design Power cooler overclocks Thermal design power - Wikipedia, the free encyclopedia

but but.. The thermal threshold of the 8350 is 80c ?

incorrect it is 62c max and if above it will result in damage

 

[sisrace]

Why because the chip will degrade going higher The threshold is only that much due to the TDP being 125W
If it was 95 w or 77 like intel you could take it to higher temps theory here is less TDP Thermal Design Power cooler overclocks Thermal design power - Wikipedia, the free encyclopedia

but but.. The thermal threshold of the 8350 is 80c ?

incorrect it is 62c max and if above it will result in damage

then coretemp and CPUZ is wrong then.. Good to know

 

[Solarstarshines]

but but.. The thermal threshold of the 8350 is 80c ?

incorrect it is 62c max and if above it will result in damage

then coretemp and CPUZ is wrong then.. Good to know

With AMD chips it is supposed to be a +10c added to temp from start to finish because it can't be recorded properly because of the architecture

Pretty lame i know it is semi close now but still off about 10c i usually shoot for max temp of 55c so i have room in case ,I wouldn't go over that because that's when you will struggle to keep it in range

 

[sisrace]

incorrect it is 62c max and if above it will result in damage

then coretemp and CPUZ is wrong then.. Good to know

With AMD chips it is supposed to be a +10c added to temp from start to finish because it can't be recorded properly because of the architecture

Pretty lame i know it is semi close now but still off about 10c i usually shoot for max temp of 55c so i have room in case ,I wouldn't go over that because that's when you will struggle to keep it in range

But if CPUz and Coretemp says that it has a max temp of 80c isnt that right?
And yeah i know they cant record the temps right but having the cpu at no more than 45c at max temp with the cooler i have is that good? or do i need to change in the bios that instead of the fan going up 0.75% per 1c to 1% or higher?

 

[Solarstarshines]

it says TJ max 90 that isn't true

like i said before the Architecture made it hard to read AMD FX-Series FX-8350 - FD8350FRW8KHK / FD8350FRHKBOX

if you are using a asus board i bet your bottom dollar it will warn you of CPU centegrade if you pass 60c

 

[madcratebuilder]

AMD reported temps are terribly inaccurate at lower temps, say below 30C but as the temps increase so does the accuracy.

Different motherboards have different reporting capabilities, my CHVFZ reports the AMD chip sensor, a ROG cpu sensor and a imbedded diode temp for socket temp, plus MB and both bridges.

AMD has not publish a official max temp. AMD support tells user not to exceed 61C. Most OC'ers well tell you not to exceed 50-55C for 24/7 operation.

I've run my FX chips in to the 70's on a few occasions and many OC'ers have run in to the 90C's for short runs.

OP, you have to understand these higher temps decrease the life span of the chip. The chip does not fry at lets say 75 or what ever, but it does shorten the life of the chip. That is why most recommend under 62C or 61C max. The quality of your motherboard effects how the chip runs at these higher temps.

 

[sisrace]

it says TJ max 90 that isn't true

like i said before the Architecture made it hard to read AMD FX-Series FX-8350 - FD8350FRW8KHK / FD8350FRHKBOX

if you are using a asus board i bet your bottom dollar it will warn you of CPU centegrade if you pass 60c

ok. Should i go into my bios and change the cpu heat warning to 60c instead?

 

[Solarstarshines]

it says TJ max 90 that isn't true

like i said before the Architecture made it hard to read AMD FX-Series FX-8350 - FD8350FRW8KHK / FD8350FRHKBOX

if you are using a asus board i bet your bottom dollar it will warn you of CPU centegrade if you pass 60c

ok. Should i go into my bios and change the cpu heat warning to 60c instead?

No you should be fine just follow what we told you and keep a eye on them and don't go above 55c on core temp or HWmonitor

 

[MagusMagnus]

I would recommend CUPID HWMonitor for the 8350 because. (There was a better/clearer, official AMD response on this but I can't find it right now. But this is the same gist of it).

So, in CUPID HWMonitor...
Temperatures > CPU = Socket temperature. <- Don't allow this to exceed 70C/158F.
AMD FX-8350 > Package = Core temperature. <- Don't allow this to exceed 62C/140F.

It's just simpler and if you keep this in mind leaves no room for mistaking one for the other.

IF you do allow the package/core temperature to consistently exceed 62F, I can and in no uncertain terms promise you that it will succumb to chronic "mini freezes" and because beyond 62F is where the actual Silicon (where the core dies are) will begin to degrade and thereby impede performance, including "glitching"; and it will be permanent damage if it happens too often and for too long. It is not pretty and it is extremely frustrating to deal with if you can't run out and replace the chip to "solve" the problem.

The other three of our systems had that happen (Intel's: one P4 and two Core2Duo's). I can't remember what the P4's maximum was, so we'll just stick with the Core2Duo's for the following. Not even for Intel can you trust the "TJMax" because those certainly were not operating at or anywhere near it (90C for the Core2Duos) they were actually each operating at a maximum of 75C (side-note: our ambient is consistently between 80F/26.6C and 110F/43.3C for half of the year, so that is definitely half of the culprit; the other half is that they were OEM builds and they ALWAYS screw you on the cooling and these two happened to also be laptops so...) and that still happened to each of them. This very thing is precisely why I forked out for the NH-D14 on this build, that should not happen this time even with this 8350... or I will simply underclock a bit to compensate (I'd rather do that than to not have an operational one at all. Don't want to deal with worrying about the potential disasters of water cooling).

Lastly, don't bother with "AMD Overdrive" for temperatures because "Starting with AMD OverDrive Ver 4.3.1.0690 instead of displaying CPU temperature, AOD will report Thermal Margin. Thermal Margin indicates how far the current operating temperature is below the maximum operating temperature of the processor" it's next to useless for that.

 

[sisrace]

I would recommend CUPID HWMonitor for the 8350 because. (There was a better/clearer, official AMD response on this but I can't find it right now. But this is the same gist of it).

So, in CUPID HWMonitor...
Temperatures > CPU = Socket temperature. <- Don't allow this to exceed 70C/158F.
AMD FX-8350 > Package = Core temperature. <- Don't allow this to exceed 62C/140F.

It's just simpler and if you keep this in mind leaves no room for mistaking one for the other.

IF you do allow the package/core temperature to consistently exceed 62F, I can and in no uncertain terms promise you that it will succumb to chronic "mini freezes" and because beyond 62F is where the actual Silicon (where the core dies are) will begin to degrade and thereby impede performance, including "glitching"; and it will be permanent damage if it happens too often and for too long. It is not pretty and it is extremely frustrating to deal with if you can't run out and replace the chip to "solve" the problem.

The other three of our systems had that happen (Intel's: one P4 and two Core2Duo's). I can't remember what the P4's maximum was, so we'll just stick with the Core2Duo's for the following. Not even for Intel can you trust the "TJMax" because those certainly were not operating at or anywhere near it (90C for the Core2Duos) they were actually each operating at a maximum of 75C (side-note: our ambient is consistently between 80F/26.6C and 110F/43.3C for half of the year, so that is definitely half of the culprit; the other half is that they were OEM builds and they ALWAYS screw you on the cooling and these two happened to also be laptops so...) and that still happened to each of them. This very thing is precisely why I forked out for the NH-D14 on this build, that should not happen this time even with this 8350... or I will simply underclock a bit to compensate (I'd rather do that than to not have an operational one at all. Don't want to deal with worrying about the potential disasters of water cooling).

Lastly, don't bother with "AMD Overdrive" for temperatures because "Starting with AMD OverDrive Ver 4.3.1.0690 instead of displaying CPU temperature, AOD will report Thermal Margin. Thermal Margin indicates how far the current operating temperature is below the maximum operating temperature of the processor" it's next to useless for that.

I REALLY hope you are not beeing serious about not letting the cpu to be running at 62F constatly Becouse otherwise i might have to move to the north pole and game outside in the snow..