It seems that some Core Ultra 9 285K processors throw WHEA errors whenever the E-core voltage is set above 1.2V.

https://www.youtube.com/watch?v=bkkzcWY39FQ

Introduction

Hello and welcome to the first CheckPoint, a new blog series for talking about specific overclocking topics in a shorter format.

Today, I want to show you this particular issue with my Core Ultra 9 285K which I ran into while preparing my SkatterBencher overclocking guide. Specifically, on this particular CPU I get WHEA errors when I set the E-core voltage over 1.2V. This happens no matter what method we use to configure the voltage and, despite the error, it doesn’t seem to hurt instability or performance.

I’ll illustrate its behavior in a second but before that I do want to mention that not all 285K CPUs have this issue. I have another sample on hand which doesn’t exhibit this problem and, based on talking with other people in the Arrow Lake overclocking scene, it seems that most don’t have this issue. Also, the Core Ultra 5 245K I overclocked in SkatterBencher #83 didn’t exhibit this problem.

I also talked about it with a couple of other people in the overclocking scene and this seems to be an issue that can also occur with retail variants, as shared by below OCN forum user, though it’s not quite clear what’s the root cause.

I spent quite a bit of time already debugging the problem and it might be related to cache or memory, though that’s just speculation at this point. What we do know is that the issue “scales” with voltage in the sense that more voltage will yield WHEA errors faster and in higher quantity.

Alright, without much further ado, let me show you how you can catch the error.

285K WHEA Demonstration

The easiest way to illustrate the behavior is by simply going into the BIOS and setting the E-core DLVR voltage to 1.25V in override mode. Then go into the operating system, open HWiNFO, and launch OCCT. Now we can run an OCCT stress test, and we see WHEA errors pop up quickly.

The WHEA errors will continue to rack up even when we sit just idly in the operating system. You can also loop Cinebench runs to illustrate the behavior.

Even if you don’t set the voltage from the BIOS, you might get errors with E-cores over 1.2V. For example, I can boot in with an override voltage of 1.195V and I can run OCCT without getting errors. However, when I adjust the voltage in the operating system to slightly above 1.2V, the WHEA errors start to emerge.

However, I can swap to another Core Ultra 9 285K and not have this issue.

Let’s now try the same test with E-core DLVR voltage set at 1.25V. As you can see there are no errors. I can even set it to well above 1.3V and there’s no errors.

285K WHEA Workaround

As I mentioned earlier in the video, unfortunately, I currently don’t have any workaround to solve the 1.2V E-core WHEA error. Together with other experts we tried to debug the issue but for now couldn’t pinpoint the root cause. If we find out more in the future, I’ll definitely share it on my blog.

For now, the only set I can take is to limit the E-core voltage to maximum 1.2V which, to be frank, isn’t that big of a deal considering it will already make the CPU hit TjMax in all-core workloads.

Outro

Alright that’s it for this CheckPoint – let me know if you have any ideas in the comments below and see you in the next one!