Question Why is one at 70? and the rest below 40?

[Im_A_Blimp]

I believe those ones at the top are all motherboard temperatures. The CPU one is at the bottom, right now my CPU is working very hard, under full load. I just cant figure out why one is 70 and the others are below 40. Any ideas??

https://imgur.com/a/GPNDdLU

View: https://imgur.com/a/GPNDdLU

 

[Karadjgne]

Because you have a 99% load on the cpu, but that doesn't mean the rest of the motherboard or case temps will get to 70°C. Case temps rarely exceed @ 40°C unless your room is excessively warm. Expect most times you'll see those temps at @ 10°C ± from what your room is. Of course there will be sensors nearer/next to the VRM's and the gpu, so those will fluctuate more widely depending on their respective loads.

My guess is the 1st reading at 72°C is VRM temp sensor, which will be high since the cpu load is high and demanding a lot of power through the VRM's. The others are things like ssd or hdd, Sata chipset, pcie hub etc.

 

[Im_A_Blimp]

Because you have a 99% load on the cpu, but that doesn't mean the rest of the motherboard or case temps will get to 70°C. Case temps rarely exceed @ 40°C unless your room is excessively warm. Expect most times you'll see those temps at @ 10°C ± from what your room is. Of course there will be sensors nearer/next to the VRM's and the gpu, so those will fluctuate more widely depending on their respective loads.

Ok, but what is 70 degrees right now, what part of the system is 70. That can't be the CPU because the CPU is on the bottom. So what is it?

 

[Karadjgne]

Edited.

 

[Im_A_Blimp]

Edited.

Ok I understand. Well given the board is a 12, 2 phase vrm, it should be able to handle this chip with no heating issues. Also the other thermals were in the 40s, low 50s, I put a fan in there to drop thermals. All the thermals dropped by 15 degrees besides the one at 70. So I'm trying to figure out exactly what that 70 is. It shouldn't be the vrm because the chip is not overclocked. Everything is running stock

 

[Karadjgne]

What does OC have to do with anything? It's a Ryzen 3900x at a full core burn, pulling maximum voltage and amperage through the VRM's. They will get warm. And stock settings can and are higher than a mild OC usually since all cpus are different Intel and amd purposely set stock power requirements higher than necessary to guarantee stability of any cpu. A mild OC will actually lower temps as you fine tune voltages.

 

[Im_A_Blimp]

What does OC have to do with anything? It's a Ryzen 3900x at a full core burn, pulling maximum voltage and amperage through the VRM's. They will get warm.

Right. I understand that. But a VRM of this size shouldn't get this warm at stock. That seems pretty high. This is the 2nd highest end board on Gigabytes X570 line

 

[Karadjgne]

Power generates heat. At that load you are pulling every bit of 140w through those VRM's, at the stock settings. Assuming you have the cooling for it, you'll be boosting like a bandit.

Drop the load, see if the temp drops. I assumed it was the VRM's based on that, but it could just as easily be the pcie hub or the Sata chipset.

 

[Im_A_Blimp]

Well first off, this CPU only has a max wattage of 105W. I don't know where you got 140 from. Anyway, I think that corsair app is garbage, because when I load the SIV from gigabytes website, there are 7 thermals, one is the CPU temperature, and one is all the other system thermals at 40 or less. So why is Corsairs iCUE reporting the CPU in the 70s, but also a system temperature on the board in the 70s? Something is a miss here. I have enclosed the SIV view as a screenshot here. Look at it compared to corsairs. Something is off. Any ideas? Because I am very confused right now

https://imgur.com/a/cqUTnoB

View: https://imgur.com/a/cqUTnoB

 

[DSzymborski]

No, the maximum wattage is not 105W. That's not what TDP means. Thermal design power isn't some rigorously calculated measure but a generalized rule of thumb for speccing out how much heat a cooling solution will need to remove.

The power threshold designed for the socket for 105W TDP processors is 142W, not 105W.

https://www.gamersnexus.net/guides/...lained-deep-dive-cooler-manufacturer-opinions

 

[Im_A_Blimp]

No, the maximum wattage is not 105W. That's not what TDP means. Thermal design power isn't some rigorously calculated measure but a generalized rule of thumb for speccing out how much heat a cooling solution will need to remove.

The power threshold designed for the socket for 105W TDP processors is 142W, not 105W.

https://www.gamersnexus.net/guides/...lained-deep-dive-cooler-manufacturer-opinions

Ok, but that doesn't explain the discrepancy with the thermals

 

[DSzymborski]

Ok, but that doesn't explain the discrepancy with the thermals

I don't see anything particularly odd.

 

[Im_A_Blimp]

I don't see anything particularly odd.

When I load the SIV from gigabytes website, there are 7 thermals, one is the CPU temperature, and one is all the other system thermals at 40 or less. So why is Corsairs iCUE reporting the CPU in the 70s, but also a system temperature on the board in the 70s? Something is a miss here. I have enclosed the SIV view as a screenshot here. Look at it compared to corsairs. SIV shows the CPU RPM as 1900. iCUE shows the CPU RPM as 1010. Something is off. Any ideas? Because I am very confused right now

 

[DSzymborski]

When I load the SIV from gigabytes website, there are 7 thermals, one is the CPU temperature, and one is all the other system thermals at 40 or less. So why is Corsairs iCUE reporting the CPU in the 70s, but also a system temperature on the board in the 70s? Something is a miss here. I have enclosed the SIV view as a screenshot here. Look at it compared to corsairs. SIV shows the CPU RPM as 1900. iCUE shows the CPU RPM as 1010. Something is off. Any ideas? Because I am very confused right now

These sensors are not universal from software-to-software. In any case, there's no screenshot that shows the 1900 RPM; that's not actually a scrollbar, that's an image of a scrollbar.

You're also assuming that all software uses an identical sample size and sample time for the data displayed. That's not how these things work.

At this time, you've presented no evidence that there's actually anything worrisome going on.

 

[Karadjgne]

Point is, there's no label on those graphs. There's an assumption that the lowest is the ONLY cpu temp, but that's just an assumption. Because of lack of knowledge regarding the exact addressing of the upper sensors, it's entirely possible that that too is a cpu temp reading hottest core. Or average cores. You simply do not know.

All I see is you've pushed a 99% cpu load, you are getting 70°C for sure on the cpu, from the lower reading. Any other readings are moot because they are all within specs. If you had one at 100°C + I'd be worried, but anything to do with power delivery can hit those temps. PCH or VRM's or Sata chipsets can easily reach 90°C + when pushed. So a 70°C is chimp change.

 

[Im_A_Blimp]

These sensors are not universal from software-to-software. In any case, there's no screenshot that shows the 1900 RPM; that's not actually a scrollbar, that's an image of a scrollbar.

You're also assuming that all software uses an identical sample size and sample time for the data displayed. That's not how these things work.

At this time, you've presented no evidence that there's actually anything worrisome going on.

These sensors are not universal from software-to-software. In any case, there's no screenshot that shows the 1900 RPM; that's not actually a scrollbar, that's an image of a scrollbar.

You're also assuming that all software uses an identical sample size and sample time for the data displayed. That's not how these things work.

At this time, you've presented no evidence that there's actually anything worrisome going on.

Sorry, not 1900. I meant 2009. Sometimes it's around 1900. But in that screenshot it shows 2009 rpm in the SIV, but when I load icue it's only showing 1100RPM. Yes, I know the Tempertures in the SIV are where they are supposed to be. But if you look at the first screenshot. There are two temperatures in the 70s. Ones the CPU. What's the other ones?

 

[Im_A_Blimp]

Point is, there's no label on those graphs. There's an assumption that the lowest is the ONLY cpu temp, but that's just an assumption. Because of lack of knowledge regarding the exact addressing of the upper sensors, it's entirely possible that that too is a cpu temp reading hottest core. Or average cores. You simply do not know.

All I see is you've pushed a 99% cpu load, you are getting 70°C for sure on the cpu, from the lower reading. Any other readings are moot because they are all within specs. If you had one at 100°C + I'd be worried, but anything to do with power delivery can hit those temps. PCH or VRM's or Sata chipsets can easily reach 90°C + when pushed. So a 70°C is chimp change.

I understand that. What I'm trying to figure out is why the first screenshot has 2 temperatures of 70 and the second screenshot has only one temperature in the 70s

 

[Zerk2012]

I understand that. What I'm trying to figure out is why the first screenshot has 2 temperatures of 70 and the second screenshot has only one temperature in the 70s

The first temp shown that says 72C should be your CPU socket.

Your actual CPU temp is listed next to the CPU.

From your first pick.

From the other it's not showing socket temp.

 

[DSzymborski]

Sorry, not 1900. I meant 2009. Sometimes it's around 1900. But in that screenshot it shows 2009 rpm in the SIV, but when I load icue it's only showing 1100RPM. Yes, I know the Tempertures in the SIV are where they are supposed to be. But if you look at the first screenshot. There are two temperatures in the 70s. Ones the CPU. What's the other ones?

Again, all these definitions aren't the same and the categories listed aren't all the same. And the sample times and sample sizes are different, and so on.

I wish you luck in your quest, but honestly, at this point, there's no real reason to think anything's wrong or the least bit odd. As far as I can tell, you've essentially invented yourself a problem to be worried about and there are other questions from people with actual problems that this is taking time from.

 

[Im_A_Blimp]

Again, all these definitions aren't the same and the categories listed aren't all the same. And the sample times and sample sizes are different, and so on.

I wish you luck in your quest, but honestly, at this point, there's no real reason to think anything's wrong or the least bit odd. As far as I can tell, you've essentially invented yourself a problem to be worried about and there are other questions from people with actual problems that this is taking time from.

Well. I don't get it. One application says the fan is spinning at 2009RPM. The other says it's 1100. Which one is it? It can't be both

 

[DSzymborski]

Well. I don't get it. One application says the fan is spinning at 2009RPM. The other says it's 1100. Which one is it? It can't be both

Both the screenshots you posted have a fan spinning at 2009 RPM.

For the umpteenth time, the way data is sampled and what sensor is identified as what and what sensors are listed are not standardized among software packages.

Do you actually have any concrete basis to think anything's wrong with your PC that you need help with? If not, I'm going to close this thread lest it become a timesink.

 

[Im_A_Blimp]

Both the screenshots you posted have a fan spinning at 2009 RPM.

For the umpteenth time, the way data is sampled and what sensor is identified as what and what sensors are listed are not standardized among software packages.

Do you actually have any concrete basis to think anything's wrong with your PC that you need help with? If not, I'm going to close this thread lest it become a timesink.

Sorry. I was referring to a different application. Here are the two applications side by side. Different fan speeds. Why??

https://imgur.com/a/qiySBV3

View: https://imgur.com/a/qiySBV3

 

[DSzymborski]

Sorry. I was referring to a different application. Here are the two applications side by side. Different fan speeds. Why??

https://imgur.com/a/qiySBV3

View: https://imgur.com/a/qiySBV3

Once again, data isn't sampled identically. 1985 RPM vs. 2009 RPM is not meaningful in any way whatsoever. And no, just because something is labeled something on one bit of software does not mean that it's not labeled differently, depending on where it's plugged in, in other software. Corsair recommends plugging the pump into the CPU fan header and assuming you followed the directions, it's not shocking at all to see that's where the pump header is being reported by other software.

I think we're done here. At this point, if you're trolling, there's no reason to keep the thread open and if you're on the level, Karadjgne and Zerk have more than answered your questions.