Question What had been advised about Intel Z170 "C"-states in overclocking guides?

[BonzaiDuck]

Just off top of your head, anyone know the answer to this? They were ASUS boards. Overclocking guides advised turning certain C-States "off", disabled or something-or other. I've still got my random bi-monthly average shutdown problem. Uncovered some advisories about Windows 10 power settings.

If anyone remembers -- I had threads in Windows, Memory and Storage, PSUs -- all a continuing month of panic about getting this system back to totally reliable, stable and forever.

I've replaced the graphics card and its PCIE eight--pin wire that had been sorely twisted into a distorted shape for cable management.

Ran the INtel CPU diagnostic for about 50 or 60 iterations -- "Pass," Pass," . . . . . "Pass,'" "Pass."

The blog about the power settings made me think that I'd changed the C-states by some advice in an overclocking guide -- when I had first built the system with the Skylake and same spec GSKILL RAM. That was for Skylake, this motherboard was originally for Skylake, BIOS-upgraded to Kaby. So I went through two perfectly-good motherboards, one an open-box purchase, the other an RMA return.

True, I'm not overclocking it, except for the memory by setting the "OC" kit to its spec and putting VCCIO/IMC to 1.15V But that was what I was doing for the Skylake after being fresh from studying the OC guides. When I bought the replacement motherboard and then tried out the RMA replacement from ASUS, I had possessed the RMA replacement board for a year before checking to see if MB as my trouble. Both boards behave the same, and so does the random shutdown occurrence and frequency.

I think it would been wise to retrieve that setting and find out what it should be.

Maybe somebody knows. I have so much other stuff to do, and this PC is important. Well -- I guess I made that plea enough times already.

I think the item was the C0 power-state. Maybe I had written it down, but I will look for it without holding my breath. One of the first things I did for the original board, and didn't do it yet for these others. It still seems sort of flakey that it could be something like this. . . .

 

[igor_kavinski]

I do understand the quest for stability but does the PC need to stay active 24/7 without fail because it is running something critical? If not, just shut it down when not in use.

If the shutdowns happen when you are not around, that might point to a power state issue, in which case disable the power saving options in BIOS.

If the shutdown/restart happens in the middle of doing something important, well, that is a big problem and I wish you best of luck in figuring that out.

Also, since you want a perfectly stable system, maybe turn off XMP and just run the RAM at stock settings.

 

[Shmee]

Are you sure this is stability related? Maybe just the computer restarting for Windows updates?

 

[BonzaiDuck]

Are you sure this is stability related? Maybe just the computer restarting for Windows updates?

Addressing both Shmee and Igor. I started another thread about this, with a title that includes the word ****.

I cannot confirm yet absolutely, but the very same symptoms -- random, occasional shutdowns and (also) unmanaged restarts -- seem to be evidence of a Windows power-management issue, on several threads, posts and other forums. The recommendations are all parallel or similar.

I just think if I'd run this thing for four years without a single twitch of instability using the Skylake processor, the Kaby Lake shouldn't be a cause of a problem with this motherboard. But there's nothing left to suspect other than Windows.

I noticed that I've posted these threads in CPUs and OC'ing. maybe they should've been put up elsewhere, but I've left a trail all over Mem and Storage, Motherboards . . . . so forth.

Anyway, I face a dilemma of old habits about building new systems. I want to test a new build for at least a couple months before it "goes live". My time right now is overwhelmed. It would otherwise be so easy to pick up a late-gen processor, motherboard and RAM, using all the same case hardware and drives. All the other parts are fresh and in my possession. But I don't have time to do it as I'd wish, although I don't have much more time for this . . ..incredible . . . trail of sweat and misery.

Sure! It has a power event ever week or so. Some could say -- one of you insinuated -- I could "live with it". But I insist that sleep and hibernate work. I insist that EIST work so the processor voltage should drop to around 0.7V when the system idles. I can't have it running at 4.5 Ghz no matter what. For that, I'd incline toward the latest-gen build. A catch-22.

I have a lot of work to do on this system, and doing it on my old backup is -- a tad more troublesome. You know -- "work done with computers" -- not "work done building and testing them".

I've got a friend across town who says he just bought three systems -- corporate asset refurbs -- for $150 each. His constant remark is "But -- it's your HOBBY!" I just don't like OEMs. I've got a tad more money than he does. I can't tell if he's laughing about this -- or what.

On my canopied patio table, I'm now building the twin to this system but with the original Skylake processor and the workstation motherboard. It may be that one of these is going to get a mobo-bundle swap-out sooner -- or later. Even that seems extravagant.

So . . . try eldercare. See how much time you have to do things the way you think they should be done. It will take me a good part of the remainder of this year to acquire parts and start testing a new-gen system. I can't be "using it" if I haven't assured myself that it's tip-top from the git-go.

Right now, that path is not something I feel eager about. And already -- poor Moms. Between the paper-work, the garden, arranging for in-home care alternatives, diaper-detail, and Moms wanting me to be in her room all the time or feeling lonely because I'm outside, in the garage, in my room -- I'm about ready to snap.

Pick any other year in my life -- this could've happened with less of a hassle. 2022: My brother died the day after new years, and I still have things to do in that regard. I'm having cataract surgery sometime this summer, so I've got to arrange respite-care for Moms.

So -- not eager to start right away with a latest-gen build. I've got older parts; I need something like this one, or -- until this gets resolved. A pile of documents to scan, telephone numbers to call, finances to manage. Oh -- I know! People will say "Well how do you have so much time for these TLTR posts?!"

I have fast fingers and I've got to vent. I can almost apologize for that.

 

[BonzaiDuck]

Anyway, if you skip the last one, I remembered from 2017 the OC-guide advice about turning off a particular C-state "report" feature in the BIOS. Since I'm running this at stock settings, it couldn't matter for the indications I mentioned about other forums show this as a Windows problem. Did you read that Windows Community thread? Suddenly windows has a "hybrid shutdown" feature? What on earth is that? And what good could it do? Maybe that's just in the mind of the "independent advisor" who mentioned it.

 

[igor_kavinski]

Suddenly windows has a "hybrid shutdown" feature? What on earth is that? And what good could it do?

Microsoft Archives

www.techrepublic.com

Not sudden. Default since Windows 8. It works so well that most of us don't even know it's there.

 

[BonzaiDuck]

Microsoft Archives

www.techrepublic.com

Not sudden. Default since Windows 8. It works so well that most of us don't even know it's there.

Sure -- that's right -- and I never heard of it either. But that point was raised by some "expert" on a Microsoft forum who wasn't even an MS employee. It seems likely or apparent that there are OS, software and driver causes behind "similar" restarts and shutdowns. I saw one guy getting a load of advice, did everything they told him, and No Cigar. He just concluded "Something is broke" as he pulled the string on a new computer.

And again -- on a Ten Forum thread where people are troubleshooting what they THINK are windows problems, a guy with the problem just solved it himself. So I solved it, too.

PCs could go to sleep or hibernate in the 1990s. I personally just never pursued perfection with those features in those days. About five years past the millennium, I gave it some attention. Over the past 5 or 10 years, I began insisting that the features had to work on every PC I owned, even if I didn't use them frequently. But the sleep states and the two essential features had nothing to do with my problem after all.