Lenovo X1E Thermals - HWInfo PROCHOT

[rx27]

Hi - 

I have recently acquired a Lenovo X1E (8750h). I am currently reviewing it's thermal performance and trying to understand some of the attributes that are displayed in HWInfo 64.


At idle (screenshot below), the system seems to be operating relatively warm ~60C. But also there are a number of Performance Limit Reaons being triggered.

[font=verdana, arial, helvetica, sans-serif]IA: PROCHOT [/font]
[font=verdana, arial, helvetica, sans-serif]IA: Electrical Design Point/Other (ICCmax,PL4,SVID,DDR RAPL) [/font]
[font=verdana, arial, helvetica, sans-serif]IA: Max Turbo Limit [/font]
[font=verdana, arial, helvetica, sans-serif]IA: Turbo Attenuation (MCT) [/font]
[font=verdana, arial, helvetica, sans-serif]GT: Max VR Voltage, ICCmax, PL4 [/font]
[font=verdana, arial, helvetica, sans-serif]GT: Fuses limit RING: Max VR Voltage, ICCmax, PL4[/font]

[font=verdana, arial, helvetica, sans-serif]I am not sure why all these are being triggered given the system is idle and temperatures are at ~60. For example, I thought PROCHOT isn't triggered until around 100C?[/font]

Interestingly, if I run ThrottleStop it does not show the PROCHOT being triggered (see attached), contradicting HWInfo. Also, as you would expect, if I disable BD PROCHOT in ThrottleStop then IA: PROCHOT in HWInfo isn't triggered. 

[font=verdana, arial, helvetica, sans-serif]Any thoughts or help with interpreting the results would be welcome. I wondering if I should repaste the CPU/GPU, or even return the unit. But before this, I would like to understand the results a little deeper. [/font]

[font=verdana, arial, helvetica, sans-serif]Thanks,[/font]

 

[Martin]

Performance limit reasons do not indicate a critical condition, but reasons why the CPU clock was reduced from the maximum value possible.
Thermal Throttling flags in HWiNFO mean that the CPU is overheating (above Tj,max) and should represent activation of PROCHOT.
I think that in your case the Bi-directional PROCHOT signal is used by some other system component (EC?), which is using other parameters (i.e. case/skin temperature) to trigger PROCHOT to reduce the performance. This is not critical though.

 

[rx27]

Performance limit reasons do not indicate a critical condition, but reasons why the CPU clock was reduced from the maximum value possible.
Thermal Throttling flags in HWiNFO mean that the CPU is overheating (above Tj,max) and should represent activation of PROCHOT.
I think that in your case the Bi-directional PROCHOT signal is used by some other system component (EC?), which is using other parameters (i.e. case/skin temperature) to trigger PROCHOT to reduce the performance. This is not critical though.

Thank you Martin for replying. 

I did some further investigation, and have resolved the problem with PROCHOT. I had been testing the machine at work, where I only had a 45W AC adapter, rather than the standard 130W one. The 45W AC adapter is enough to charge the laptop. However, when using it, it seems the PROCHOT event is frequently triggered by something external to the CPU, which I can now suggest is related to the AC power availabilty. When I use the 130W AC adapter PROCHOT is not triggered. 

I am still seeing the following events being triggered - 

[font=verdana, arial, helvetica, sans-serif]IA: Turbo Attenuation (MCT) [/font]
[font=verdana, arial, helvetica, sans-serif]GT: Max VR Voltage, ICCmax, PL4 [/font]
[font=verdana, arial, helvetica, sans-serif]GT: Fuses limit RING: Max VR Voltage, ICCmax, PL4[/font]

[font=verdana, arial, helvetica, sans-serif]Are you able to provide me some explanation for these?[/font]

[font=verdana, arial, helvetica, sans-serif]Thanks,[/font]

 

[Martin]

Those can have several reasons.
IA: Turbo Attenuation (multi-core turbo demotion) cause of frequency clipping - e.g. when frequently transitioning between 1C and 2C turbo, the latter ratio limit will be used for both.
GT/RING: Max VR Voltage, ICCmax, PL4 - maximum designed voltage regulator (VR) voltage reached for the GT/RING VRs
Also, some CPUs might due to a bug in the CPU report some invalid reasons for the RING.