Maximus X Hero Wifi AC Hard Freeze at Idle

[akw143]

Hey all,

I've been dealing with an intermittent issue for the past couple of months, sometimes I think I've fixed it, but so far it's always come back over about the past 3 months.

Basically, my PC runs perfectly 99% of the time, games very well, crushes benchmarks, is stable as a rock (except for in certain scenarios - which we'll get to), and all around great.

However, I run into issues at or near idle with the CPU. I get a complete hard freeze after about 10-30 minutes of idle or light loads (YouTube, browsing the web, chatting on Discord, etc.). Hard freeze meaning a complete lock up, no BSOD, no warnings or error messages, just a complete lock up including KB+M, even the reset button doesn't work - I have to hold down the power button to shut down completely. It will not unfreeze itself or automatically reboot no matter how long I let it sit. Event Log gives me an abnormal reboot error afterwards, but nothing unusual up to and including the freeze itself. This does NOT happen at all under any significant load, whether it's gaming, rendering, benchmarking, etc. - the system is perfectly stable at all times under load. I have completely cleared all overclocks and other UEFI customizations and gone back to bone stock and the issue persists. It is not the easiest thing in the world to effectively reproduce - I kind of just have to do boring light load applications until it decides to rear its head. Like I said, most times within 10-30 minutes.

I thought maybe I had my Vcore set too low, adjusted that, no good. Disabled C-states, same thing. The only stupid fix I've found to be 100% effective (so far) is setting the power management options in Windows to never allow the CPU below 100% state, which is pretty annoying since I'm pushing constant max voltage and cannot take advantage of power saving features as well as potentially shortening the life of my CPU.

Specs for reference:

Maximus X Hero (Wifi AC)
i7 8700k
32GB Corsair Vengeance RGB @ 3600 XMP
EVGA 1080ti SC Black Edition Gaming

Those are the important bits, the whole thing is cooled by an EK-FG liquid cooling loop.

My latest attempt at a fix was to upgrade to BIOS 1704, since it had mention of CPU Microcode updates, but the problem arguably got worse or at best stayed the same. HOWEVER, I noticed that the issue *may* have started around the time we got BIOS 1602, the one with 9x00 processor support. With that in mind, I actually rolled back to BIOS 1401, the one just previous to 1602 and so far I *think* the problem might be solved.

If anyone else has had this issue and come up with a different solution, let me know. Any other advice is appreciated. As I said, this issue is not the easiest to reproduce, so I'm not 100% convinced that the 1401 downgrade has it fixed, but I'm confident.

I will report back in a couple of days as well in case anyone comes here with the same issue.

Thanks in advance for any advice.

[NemesisChild]

Welcome to the forum!

Is anything else overclocked besides the memory?
Have you tried testing at default bios settings over a long period of time?
Have you tried clearing the CMOS?
Have you run Memtest?
Have you tried a fresh installion of Windows?

[akw143]

Welcome to the forum!

Is anything else overclocked besides the memory?
Have you tried testing at default bios settings over a long period of time?
Have you tried clearing the CMOS?
Have you run Memtest?
Have you tried a fresh installion of Windows?

CPU is/was - tried default clocks including multicore enhancement off, same result. Same for GPU.

Default BIOS settings were as bad or worse pretty much immediately.

Cleared CMOS as well as reflashed BIOS versions, which should have the same effect.

Ran memtest, no issues.

Just reinstalled Windows last night, issue popped back up almost immediately. That's what led me to try the 1401 downgrade.

[NemesisChild]

Set your Dram voltage to 1.35v and see what it shows in the bios after a reboot.

Is your memory a single matched set? 2x16GB or 4x8GB set?

Also please note the make, model, and age of your PSU.

[akw143]

Set your Dram voltage to 1.35v and see what it shows in the bios after a reboot.

Is your memory a single matched set? 2x16GB or 4x8GB set?

Also please note the make, model, and age of your PSU.

DRAM always runs at 1.35V, since I built the machine. I'll check what BIOS reports it as in a few minutes. E: BIOS reports 1.344V. Single matched set of 4x8GB.

PSU is less than a year old, 850W EVGA Supernova G3 80+ Gold.

[akw143]

Just a quick update, it's been about 24 hours now with no hard freezes after downgrading to BIOS 1401. I've also done an unnecessary amount of idle workloads in the past 24 hours, trying to illicit the issue and haven't seen it pop back up yet. Not a guaranteed fix yet, but it seems to be much, much better than before.

[akw143]

Update #2 - just had another hard freeze so it appears that although the issue is less frequent after the downgrade, it still persists. However, a new thought process came to light. During many (if not all - can't recall exactly) of the crashes, I had HWInfo64 open. It's a pretty common practice for me to run it on my second monitor to watch temperatures and other info. I almost always have it running. I also always tick the "don't warn me again" box about the Asus EC sensor conflict issue when I install the software because I never thought it would be a big deal. In the past 24 hours or so without crashes, I hadn't had HWInfo64 open much, if at all, because I was mostly idle testing and thought that the problem had fixed itself and so felt no need to monitor the system that closely. Anyway, I've disabled the EC sensor monitoring on HWInfo64 and will see if maybe that was causing my issue to begin with.

Will update again if the problem reappears.

[akw143]

Had another crash within minutes of the last one with HWInfo64 open, however with Asus EC monitoring disabled. Going to try running for a while without HWInfo64 running and see if maybe it's just causing havok with my motherboard for some reason. Hadn't really considered that before.

[akw143]

HWInfo is not the cause, had another crash about 30 minutes later without it running. Crashes are becoming more frequent so I've given up disabled speedshift and speedstep at the moment so I just idle at base clock. I have yet to have a crash with those settings disabled, I'm beginning to think my chip is just dying after the delid for some reason and will probably upgrade to a 9900k soon and hope the problem goes away.

[Nate152]

Hi akw143

Hard freezing alot of times is ram related, you could try with 1 or 2 sticks or another ram kit if you have one.

Post us a screenshot of HWinfo, maybe it will tell us something.