Issue with temperature sensor in the motherboard

[Reynolds8]

I have mounted the following workstation:

 

CPU: 2X AMD EPYC 7573X

GPU: NVIDIA RTX A4000

Motherboard: GIGABYTE MZ270-HB0 REV 40

RAM: KINGSTON DDR4-3200 16X16 GB

SSD: KINGSTON KC3000

Power supply: CORSAIR RM1000i

CPU fans: NOCTUA NH-U14S TR4-SP3

System fans: 3X ARCTIC F12, 3X ARCTIC F14

 

Most of the use is to perform Computational Fluid Dynamics simulations. I have installed HWiNFO64, and I have performed some 1h-tests to check all temperatures and how the control system of the fans performs.  The tests consists of performing simulations that use all 64 cores and around 125 GB of RAM. HWiNFO is turned on just after the simulations starts, and 1 minutes after the simulation is finished.

 

The results are attached in two screenshots (the colums are actual-minimum-maximum-average temperatures). The conclusions that I extract after following closely the 1h test are as follows:

 

- The CPUs do not overheat, and the fans seem to be well calibrated for this purpose.

- No issues with the GPU nor the SSD.

- All sensors of the motherboard seem fine except from the VR_P0_TEMP

 

Regarding the problematic sensor, is it critical? Where is it located? I have done some research on the Internet and I have found nothing. I have to say that the room is hot (around 25 celsius), but in any case I do not know how to proceed. That temperature does not seem to be dependent on the fan speed. Initially it is around 60 degrees (when there is no simulation running), and when the simulation starts, it increases rapidly until 90-95 degrees )which seems to be a steady-state temperature). In addition to the fact that 95 degrees seems to be a high value for a non-expert, I am also worried that something can happen if I leave a simulation running, for example, for two days (instead of just one hour).

 

Thank you in advance

 

 

[Blue4130]

29 minutes ago, Reynolds8 said:

I have mounted the following workstation:

 

CPU: 2X AMD EPYC 7573X

GPU: NVIDIA RTX A4000

Motherboard: GIGABYTE MZ270-HB0 REV 40

RAM: KINGSTON DDR4-3200 16X16 GB

SSD: KINGSTON KC3000

Power supply: CORSAIR RM1000i

CPU fans: NOCTUA NH-U14S TR4-SP3

System fans: 3X ARCTIC F12, 3X ARCTIC F14

 

Most of the use is to perform Computational Fluid Dynamics simulations. I have installed HWiNFO64, and I have performed some 1h-tests to check all temperatures and how the control system of the fans performs.  The tests consists of performing simulations that use all 64 cores and around 125 GB of RAM. HWiNFO is turned on just after the simulations starts, and 1 minutes after the simulation is finished.

 

The results are attached in two screenshots (the colums are actual-minimum-maximum-average temperatures). The conclusions that I extract after following closely the 1h test are as follows:

 

- The CPUs do not overheat, and the fans seem to be well calibrated for this purpose.

- No issues with the GPU nor the SSD.

- All sensors of the motherboard seem fine except from the VR_P0_TEMP

 

Regarding the problematic sensor, is it critical? Where is it located? I have done some research on the Internet and I have found nothing. I have to say that the room is hot (around 25 celsius), but in any case I do not know how to proceed. That temperature does not seem to be dependent on the fan speed. Initially it is around 60 degrees (when there is no simulation running), and when the simulation starts, it increases rapidly until 90-95 degrees )which seems to be a steady-state temperature). In addition to the fact that 95 degrees seems to be a high value for a non-expert, I am also worried that something can happen if I leave a simulation running, for example, for two days (instead of just one hour).

 

Thank you in advance

 

 

 

 

How do you know that it is problematic? It's Voltage Regulator, could very well run that hot.

[Reynolds8]

4 minutes ago, Blue4130 said:

How do you know that it is problematic? It's Voltage Regulator, could very well run that hot.

I did not explain myself well. 

 

I do not know if it is problematic or not because (before your message) I did not know the meaning of that sensor. That is indeed the question. Should I worry about that? 

[DreamCat04]

1 hour ago, Reynolds8 said:

I did not explain myself well. 

 

I do not know if it is problematic or not because (before your message) I did not know the meaning of that sensor. That is indeed the question. Should I worry about that? 

95°C on voltage regulators is pretty hot, but it's well within their spec. Most mostfets(the part that gets the hottest in a VRM) are rated for 120°C or more. I'd try having a fan blow at the voltage regulators and doing your test again

[Reynolds8]

1 hour ago, DreamCat04 said:

95°C on voltage regulators is pretty hot, but it's well within their spec. Most mostfets(the part that gets the hottest in a VRM) are rated for 120°C or more. I'd try having a fan blow at the voltage regulators and doing your test again

Thank you for your recommendation. I will try to do that.