XTU: Applying *any* change causes cold boot instabilities

[TimurB]

Hello.

On my undervolted Z790 + 13900K system I just reproduced over two dozens instant cold-boot crashes (no BSOD, sometimes clearing BIOS settings) starting a Cinebench 23 benchmark loop at realtime priority. This seems to be caused by XTU.

The instability is not (!) based on my (BIOS) CPU settings, but on Intel XTU changing anything about them, even when slower values are used or when values are just set back and forth. I compared CPU values via HWinfo before and after hitting Apply in XTU after going back and forth on a value, but could not detect any differences. So whatever XTU does is rather hidden.

Based on the severity of the instabilities I assume something about voltages being messed up despite XTU claiming not to touch those when a VF curve is used in BIOS. It does not change the VF points, though, because these are still listed as unchanged in HWinfo. 

Doing a clean installation of an older XTU version did not help. I do not see voltage differences via HWinfo between before vs. after applying XTU changes. The cold reboot only happens using CB23 at Realtime priority, not Normal or High.

(PS: I only use XTU to workaround a BIOS bug where AVX ratios are set before Core ratios after extended shutdown and thus need two boot-ups every morning to be properly set based on non stock Core ratios.)

[TimurB]

I thought this might be Ring related. There are two different Ring ratio curves for the same min:8 and max:50, one more aggressive than the other. My speculation was that XTU might activate the more aggressive curve.

But now I manually switched to the more aggressive curve and instead of instability it just lowers the CB23 scores due to more of the power limit being used up by the Ring then. Was worth a try.

[Mike_Intel]

Hello TimurB,

Thank you for posting in Intel community Forum.

For us to further check this, please help provide the following details:

1. If you are running on default settings, are you having stability issues with your system?
2. Please help generate the SSU logs of your system. Kindly download the utility here:

https://www.intel.com/content/www/us/en/download/18377/intel-system-support-utility-for-windows.html

If you have questions, please let us know. Thank you.

Best regards,

Michael L.

Intel Customer Support Technician

[TimurB]

Hello Michael,

thanks for the quick reply.

> If you are running on default settings, are you having stability issues with your system?

Surprisingly, the answer is YES. Additionally to BIOS defaults I even set PL2 253 W (8s), PL1 125 W, TVB 70°C, ICCmax 307 A and AC = DC = 90, because defaults set these differently. Applying settings in XTU (like setting AVX offset from 5 to 6) still causes an instant cold boot on the first loop of Cinebench 23 (but only when run at REALTIME priority).

To be clear, none of these BIOS settings or my own cause instabilities just from running Intel XTU. I have to specifically apply (seemingly) any setting, even when the setting is slower or when I set something back and forth again (hitting Apply both times). Using CB23 in realtime priority is just a way to reproduce the instability, as I had these kind of cold boots happen before in other situations and just couldn't pinpoint them to XTU then.

I write "cold boot", because the corresponding crash causes a full "PSU relay click-clack boot with BIOS boot failure message", sometimes even resetting the BIOS to full factory settings similar to Clear CMOS (Easy UI).

> Please help generate the SSU logs of your system. Kindly download the utility here:

I created the report and saved it as a file. The "Submit" button does not work, though (servicerequestmgmt.intel.com’s server IP address could not be found). From what I can tell there is no compromising data in the text-file, so I upload that one here as attachment.

Best regards,

Timur

[TimurB]

I found out that stopping or restarting the XTUOCDriverService after applying changes in XTU fixes the cold boot instabilities (until the next change is applied).

[Mike_Intel]

Hello TimurB,

Thank you for the quick reply.

Before we proceed, let me just set the expectation that as much as we want to assist you, we cannot guide you on the settings for overclocking because overclocking may void the Warranty of the processor. If the processor have issues on default settings, we can provide some recommendations for that.

As for overclocking, let's wait if other forum user have similar set up and they may help you with the settings.

If you have questions, please let us know. Thank you.

Best regards,

Michael L.

Intel Customer Support Technician

[TimurB]

Hello Michael,

since this problem happens at stock settings and regardless of what is set either in BIOS or Intel XTU this is not about overclocking. It seems to be a software bug of the Intel XTU service process.

The "Report an Issue" button in XTU forwarded me to this "Processors" forum, but if there is a more appropriate place for reporting XTU issues then please let me know.

Thanks and regards,

Timur

[Mike_Intel]

Hello TimurB,

Thank you for the quick reply.

Can you try to reinstall the XTU? Please follow the steps in this link:

https://www.intel.com/content/www/us/en/support/articles/000032459/processors/processor-utilities-and-programs.html

If you have questions, please let us know. Thank you.

Best regards,

Michael L.

Intel Customer Support Technician

[TimurB]

Hello Michael,

neither a Reinstallation nor trying an older version fixes this. I followed the instructions you provided.

The only thing that helps is to stop or restart XTUOCDeviceService after applying changes in XTU.

Best regards,

Timur

[Mike_Intel]

Hello TimurB,

Thank you for the quick reply.

Let's try additional test on this issue. Please try to open the link below and try the settings in the BIOS.

https://community.intel.com/t5/Processors/Updated-Guidance-RE-Reports-of-13th-14th-Gen-Unlocked-Desktop/m-p/1594553

If you are still having the same issue, try to uninstall the XTU and try the settings again and let us know.

If you have questions, please let us know. Thank you.

Best regards,

Michael L.

Intel Customer Support Technician

[TimurB]

I already tried the settings listed under the link you provided, except for enabling CEP. I now repeated the settings and also enabled CEP and XTU's XTUOCDriverService still causes the hard cold boot crashes.

[Mike_Intel]

Hello TimurB,

Thank you for the quick reply.

Can you run the Processor Diagnostic tool to check the processor. Kindly check the link below:

https://www.intel.com/content/www/us/en/support/articles/000005567/processors.html

If you have questions, please let us know. Thank you.

Best regards,

Michael L.

Intel Customer Support Technician

[TimurB]

Tested using PL1=125 W, PL2=253 W, ICCmax=307 A, AC=DC=90, JEDEC DRAM 4800 MT

--- IPDT64 - Revision: 4.1.9.41
--- IPDT64 - Start Time: 10.06.2024 11:12:50

----------------------------------------------
-- Testing
----------------------------------------------
CPU 1 - Genuine Intel - Pass.
CPU 1 - BrandString - Pass.
CPU 1 - Cache - Pass.
CPU 1 - MMXSSE - Pass.
CPU 1 - IMC - Pass.
CPU 1 - Prime Number - Pass.
CPU 1 - Floating Point - Pass.
CPU 1 - Math - Pass.
CPU 1 - GPUStressW - Pass.
CPU 1 - CPULoad - Pass.
CPU 1 - CPUFreq - Pass.

IPDT64 Passed
--- IPDT64 - Revision: 4.1.9.41
--- IPDT64 - End Time: 10.06.2024 11:16:37

----------------------------------------------
PASS

[Mike_Intel]

Hello TimurB,

Thank you for the information provided.  

We will do further research on this matter and post the response on this thread once it is available.

Have a fantastic day, and thank you very much for your patience and understanding!

Best regards,

Michael L.

Intel Customer Support Technician

[TimurB]

Hello Michael,

thanks for looking into this.

Today I did a clean Windows 11 installation (not just Reset) and the issue persists afterwards with everything being installed from scratch. I also did a Clear CMOS to get rid of any possible BIOS settings corruption.

So this really seems like a XTU service software bug. My current workaround is to automatically stop the XTU service 1 minute after it has been started via Task Scheduler.

Best regards,

Timur

[Mike_Intel]

Hello TimurB,

Thank you for patiently waiting for our update.

Base on the data that you provided here are our findings.

Running such demanding process as real time priority is not a good idea.

Please refer to the Dev blogger link: https://devblogs.microsoft.com/oldnewthing/20100610-00/?p=13753

Information from Windows insider book:

We think it might be an expected behavior because as mentioned in above text and article that benchmark will take over entire CPU time and maybe in some scenarios can cause an OS crash, because of no clock time for other processes.

Task will be done anyway, and process allocation will be done probably the same way.

We suggest to stick to whatever Windows set and if you really want to change priority, then High is max I'd set to keep OS running.

And set CPU in BIOS to Intel values.

If you have questions, please let us know. Thank you.

Best regards,

Michael L.

Intel Customer Support Technician

[TimurB]

Hello Michael,

thanks for looking into this and getting back to me.

But running 10 seconds of a single CB23 loop at Realtime does not cause OS crashes, it causes freezing for the duration of the run. And the whole thing is just a way to 100% reproduce the XTU service (!) causing hard crashes (including PSU relay clicking like a power-off/on cycle).

This is emphasized by the workaround to stop or restart the XTU service after applying any changes in the XTU application. This solves the problem. In the end it very much looks like a software bug of the XTU service!

Best regards,

Timur

[Mike_Intel]

Hello TimurB,

Thank you for the update.

We understand your side of it however, as per our Engineers, this is an expected behavior base on the document that we shared.

If you have questions, please let us know. Thank you.

Best regards,

Michael L.

Intel Customer Support Technician

[TimurB]

Hello Michael,

then your engineers are not interested in getting bug reports for their software, including means to reproduce an error. That's unfortunate.

Timur

[TimurB]

I just reproduced the same PC crashing behavior running CB23 at "High" priority. Also resulted in a "BIOS failure detected", which sometimes does and sometimes does not happen.

It's not unlikely that it would also be reproducible running CB23 at "Normal" priority, it just becomes harder and harder then, because obviously this is an edge case that doesn't happen under all load scenarios. Using "Realtime" priority is just a way of reproducing it easier (like 100%) instead of waiting for sporadic crashes.

And to emphasize this again: Running CB23 + XTU service at "Realtime" priority does *not* cause crashes if no changes were applied in XTU beforehand or if the XTU service was restarted after applying said changes!

[Mike_Intel]

Hello TimurB,

Thank you for the update.

I will update our Engineers about this again and post the response on this thread once it is available.

Thank you very much for your patience and understanding!

Best regards,

Michael L.

Intel Customer Support Technician