I'm overcocking my CPU, should I keep an eye on this?

[podkall]

Basicaly I can see that the TMPIN0 is getting pretty much same temps as my CPU,

what is this tempature for and should I be looking over so it doesn't overheat?

If yes what temp is max without damaging?

 

[Levent]

Sensor names do not correlate to same sensor on every machine. You have to identify which component's temperature it is showing you then find out the temperature range of that component.

[podkall]

4 minutes ago, Levent said:

Sensor names do not correlate to same sensor on every machine. You have to identify which component's temperature it is showing you then find out the temperature range of that component.

Well it's something on motherboard that is related to CPU, isn't that enough info?

[Levent]

1 minute ago, podkall said:

Well it's something on motherboard that is related to CPU, isn't that enough info?

How did you come up with that? If you dont know which components it is reading temperature of, that reading is still useless. 

[podkall]

Just now, Levent said:

How did you come up with that? If you dont know which components it is reading temperature of, that reading is still useless. 

You don't see the picture?

There is a picture of a motherboard, the only thing in my PC which is ASUS branded, under wtich you see voltages and tempatures.

So these are sensors of the motherboard, but I am asking, what motherboard component heats up when CPU is stressed and should I watch if it goes overboard?

[germgoatz]

he is making his cpus cock bigger

[manikyath]

sensor names can be quite cryptic.. but given that the sensor largely follows the temperature of your cpu, i assume it's either the motherboard relaying cpu temp, or it's a temp sensor underneath the socket.

 

as for in the context of cpu overclocking.. i suppose your biggest worry is the processor itself? (presuming that most motherboards these days have stupidly beefy VRM designs...)

 

EDIT: i just noticed that it's an ASUS motherboard.. asus loves to dot random temp sensors around the motherboard.. my old Z97 sabertooth mk2 has temp sensors reporting all sorts of things that make no sense what so ever.. luckily it seems that with my current 'ROG strix X570-e gaming' they seem to have gone a bit smarter on sensor placement, and a bit dumber on naming scheme...

[Levent]

1 minute ago, podkall said:

You don't see the picture?

There is a picture of a motherboard, the only thing in my PC which is ASUS branded, under wtich you see voltages and tempatures.

So these are sensors of the motherboard, but I am asking, what motherboard component heats up when CPU is stressed and should I watch if it goes overboard?

The only picture I see is this one. It doesnt matter what it says, as long as you dont know which component those sensors are placed near, those readouts dont mean anything. VRMs will heat up when CPU is using more power.

[podkall]

1 minute ago, manikyath said:

sensor names can be quite cryptic.. but given that the sensor largely follows the temperature of your cpu, i assume it's either the motherboard relaying cpu temp, or it's a temp sensor underneath the socket.

 

as for in the context of cpu overclocking.. i suppose your biggest worry is the processor itself? (presuming that most motherboards these days have stupidly beefy VRM designs...)

It's not exactly the same temps, so it's probably something under the cpu.

 

2 minutes ago, Levent said:

VRMs will heat up when CPU is using more power.

Is there any dangerous level of tempature? My CPU seems to have infinite potential when it comes to clock speed...

 

[manikyath]

Just now, podkall said:

Is there any dangerous level of tempature? My CPU seems to have infinite potential when it comes to clock speed...

there's probably a cpu VRM temperature sensor somewhere.. may even be two.. they're probably also labeled as "VRM temperature" or something like that.

 

as for safe temperatures.. VRM's can take a lot, and it depends on how it's measured...

the long answer:

 

[podkall]

2 minutes ago, manikyath said:

there's probably a cpu VRM temperature sensor somewhere.. may even be two.. they're probably also labeled as "VRM temperature" or something like that.

 

as for safe temperatures.. VRM's can take a lot, and it depends on how it's measured...

the long answer:

 

Does the motherboard force shutdown or restart or thermal throttle if VRM gets too hot?

 

Also after increasing clock, Firefox worked, but opening Cinebench made my pc go into blackscreen, so than I upped the voltage, is that a right move as long as my clock speeds (and voltage of course) aren't ridiculous?

[manikyath]

2 minutes ago, podkall said:

Does the motherboard force shutdown or restart or thermal throttle if VRM gets too hot?

 

Also after increasing clock, Firefox worked, but opening Cinebench made my pc go into blackscreen, so than I upped the voltage, is that a right move as long as my clock speeds (and voltage of course) aren't ridiculous?

i'm not by any means an expert on overclocking.. but generally i can say this:

- firefox isnt a good metric to decide if an overclock is stabile. last time i bothered to look into it, prime95 was the way to check cpu stability. essentially you want a 'worst case scenario' workload that is many times more likely to cause a crash than your daily workload is.

- more voltage means more wear on the cpu. as for which voltages are problematic for which cpu.. i'm not the guy to ask...

[podkall]

1 minute ago, manikyath said:

i'm not by any means an expert on overclocking.. but generally i can say this:

- firefox isnt a good metric to decide if an overclock is stabile. last time i bothered to look into it, prime95 was the way to check cpu stability. essentially you want a 'worst case scenario' workload that is many times more likely to cause a crash than your daily workload is.

- more voltage means more wear on the cpu. as for which voltages are problematic for which cpu.. i'm not the guy to ask...

Thanks for the insight

 

@Levent  I just blackscreened again, is it because I don't increase the voltage enough? I'm below 1,3V

[metaleggman]

I'm no expert at overclocking either, but if you feel like your processor has infinite potential, then you're likely not overclocking it right.  Whenever something has seemed like that in the past for me, I ended up finding out it was too good to be true.  And from the sounds of it, whatever settings you're using sound particularly unstable.

 

It looks like you have a Ryzen 3100, so you're best off using something like 1usmus's Clock Tuner for Ryzen, and then using OCCT to stress test for errors.  Because AVX workloads aren't common for most consumers, 1usmus's overclocks can be a little generous in my experience, as I think his tool only uses an SSE workload.  For me, I was able to tweak the voltage and clocks to where AVX1 workloads were pretty much error free on both my overvolted and undervolted OC (1usmus allows you to do both, and you probably should).

As to the sensors, if they're random motherboard temps, I wouldn't worry, assuming you have enough cases fans.  Some overclocking focused hardware (think those ridiculous motherboards, or EVGA K1NG Pin cards with a dozen ICX sensors) will include sensors solely for the end-user's use, but I'm guessing Asus installed those sensors so the BIOS can keep an eye on them.  I haven't heard anyone informing would-be overclocks to keep an eye on your motherboard temps, since case fans should keep them cool enough.  I'd also guess the motherboard would react in some way if things got too hot, but again, without knowing what the temps are for, or why Asus placed them there, we really don't know.  As was mentioned, one could be for the VRMs, which will get hot.  I'm guessing one is likely the chipset, which would need a temp sensor to power its fan.

In addition to knowing what CPU you have, I'm curious what motherboard you have.  Not all motherboards have robust enough power management for pushing a CPU to its realistic limit.

[Mister Woof]

Try HWInfo64 instead of hwmonitor