On a Intel Q67 mobo(DQ67SW), i5-2500k cpu, BIOS shows the PCH temperature is 0.
It is strange, it wasnt 0 before, I think it was 44c before. Now 4 out of 5 times entering BIOS setup, it is 0. Also the Intel ME menu button is gone at setup. I did the Intel MEBX header reset, it does not make a difference.
The system is experiencing random shutdown, usually within 30mins after win10 booted up.
I suspect it is related to the PCH thermal such that if BIOS shows PCH temperature 0, after win10 booted up, it will shut down within 30 mins. However, if BIOS PCH temp shows ~44c(regular value), it does not shutdown after win10 booted up, I can run aida64 stability test overnite and it did not shutdown. It seems to be the correlation although not 100% sure.
Also when BIOS PCH temp is 0, aida64 sensor does not show PCH temperature entry. Intel management engine interface does not show up at Device Manager either, it says the device is NOT connected, code 45. I turn on the hidden device at DM, looks like the driver was re-installed automatically by microsoft during the Win10 version 1809 update. Dont know if it is related.
Anyone has any clue what is happening here?
Thanks in advance.
I would say that, based upon your description, the PCH component is in the process of failing. It is intermittent now, but will fail completely sometime soon. This board is 8 years old, so it has passed the MTBF for many components.
When it shuts down, it just power off without any warning. Is that the typical failure signature?
It has BIOS Version 0068 - SWQ6710H.86A.0068.2017.0601.1423 on it now, is it worth trying update to latest BIOS Version 0070 - SWQ6710H.86A. 0070.2018.1024.1352 or it might not boot win10 if update the BIOS under this condition?
First of all, BIOS releases 69 and 70 just deliver updated processor microcode that address certain Spectre sideband vulnerabilities, so installing these will not address your issues - but you should be installing them anyway.
This is a possible failure scenario that you describe, but the scenarios you previously described wherein the Management Engine intermittently does not operate are also failure scenarios. A new thought: failing memory can also cause similar scenarios.
Windows 10 is not supported on this motherboard or with this processor. It may run, but not all hardware will be properly supported.
The issue has been resolved.
The above symptons occurred like 4 out of 5 powercycles/reboots, and then to the point it always happens after every powercycle/reboot.
I have flashed BIOS to latest v70(was v68), resetted Intel MEBX header, turned BIOS to config mode and resetted AMT settings, all didnt help.
I tried to reload ME FW using the utility, but it errored out "Error 8193: Intel ME Interface : Cannot locate ME device driver". That is understood, because the IME interface driver was not even running as the IME HW was reported as not connected.
Then I found that some other users experienced similar shutdown, it is not within 30 mins, it is EXACTLY 30 mins and even sitting in BIOS.
I tried that and indeed the case, and is due to corrupted ME FW or ME FW not running. It is a known scenario.
Then I forgot that I didnt do one thing => load BIOS defaults. After I did that, I immediately got back PCH temp reading in BIOS.
Win10 DM shows IME interface is working. I immediately reload ME FW, this time reloading works.
Rebooted back into BIOS, "Intel ME" menu reappeared, everything looks fine.
Somehow I think the BIOS settings together with "maybe" corrupted ME FW caused ME FW to not running. Without the latter running, all above issues occurred. After I reloaded BIOS defaults settings, it cleared thing up and able to reload ME FW.
What caused that is unknown, CMOS battery seems good. The BIOS setting values displayed were all good too. However, I think the BIOS(AMI?) only shows what's relevant and hides things not detected. So it wont show the problem entries.
P.S. there is a guy at Win-Raid forum, having very similar issue with a lenovo Z50-70 laptop, hope he sees this and may find this useful.