Skylake hard locks seeminly when IDLE / CStates ON

[GGlyn]

Myself and others have been struggling with a different hard lock problem with Skylake and we have ran out of options but to assume it is a CPU fault/issue. Does anyone have any comments or feedback?

The long thread is over here, but I will summarise. http://www.tomshardware.co.uk/forum/id-2830772/skylake-build-randomly-freezing-crashing/page-7.html New Skylake Build Randomly Freezing/Crashing - Page 7 - windows 10 - Windows 10

Our systems hard lock randomly but appears on the most part to be when IDLE or next to IDLE. The hardlock can come when playing video, browsing (reading) or mostly when not interacting at all. Some (like myself) noted that they seem to occur more often if you are doing a large amount of disk IO (partition move for example or copying 40+ GB of data disk to disk).

Hard locks happen in either Windows (7,8,10) or Linux.

With CStates on, the crashes are often and usually within 10 minutes of IDLE. With CStates disabled completely most find that the system remains "mainly" stable for days. Generally, no one has reported locks when gaming or under high CPU load (so I do not think this is the Prime95 example but then, since compilers / drivers use the latest instructions (according to CPUID) I cant be sure).

The symptoms happen on many different manufacturers motherboards, RAM types and setups. We have successfully removed or replaced all components and the only change that seems to make a difference is CPU for a few. CStates off in the BIOS is absolutely a guarenteed way to stabilize significantly but not fix.

I changed CPU (after trying everything else) and find it now mostly stable with CStates ON. I must be clear that my system is 100% identical, CPU # 1 crashes all the time and must have CStates off, CPU # 2 does not crash so often even with CStates ON. One may argue this is a problem with my PC, a fault, but with so many people having the exact same issues I cant believe it is a defect. And with Cstates/SpeedStep so specifically (a known new part of Skylakes architecture), I remain convinced that this is something real.

Others were not so lucky, a replacement CPU did not make the situation better. Batches from different factories as well etc etc

My new CPU has hard locked but not when IDLE like the last one (not yet anyway). Mine can lock when under full load, video transcoding, but not more than 1 in 10 sessions. Cstates ON seems equally stable on this CPU as OFF.

Thanks for any suggestions you can give.

[LMonk]

After your words i think better to start from motheboard. I noticed that your motherboard is pretty expensive, what do you think about for example Asrock fatality z170 k4 or asrock z170 extreme 4? Quad sli/crossfire is overkill for me.

[CCris2]

Only mainboard from Asrock that has similar power phases like mine is Asrock Z170 Extreme 6 . Asrock fatality z170 k4 and Asrock z170 extreme 4 have less CPU phases.

[LMonk]

I decide to do one last thing. I disabled all c-states and set my cpu to work always on turbo. Wanna see what will happen on such settings. Also i did such settings to digi+vrm http://puu.sh/pBcX4/8a70d4596c.jpg http://puu.sh/pBcX4/8a70d4596c.jpg

Also i moved my system parition on ssd intel 535 It is blazing fast now, dunno how i live without ssd earlier. I hope this will do my system more stable.

[CCris2]

I think that image is wrong

[LMonk]

Ahaha, lol. So sleepy! Fixed it.

[CKowa1]

Hello. I also have troubles with random freezing. Here's my spec:

Mobo: MSI Z170A Gaming Pro Carbon with old bios ver 1.00

Cpu: i6700K @ 4.4Ghz

Ram: One stick of Corsair DDR4, 8GB, 2400MHz, CL14, Vengeance LPX RED (CMK8GX4M1A2400C14R)

Psu: 650W Chieftec

SSD Kingstone V300 60GB

OS: Windows 10 Pro 64bit

I use this rig from exactly 14.05.2016. The thing is, that few days ago I sold my Nvidia MSI GTX970 Gaming 4G, and went to onboard GPU with dual monitor setup, and after that my freezing problem has begun. Before that I was playing GTAV every day and few other games with no problem at all!

[CCris2]

Hi @Makabra

I see you have overclocked CPU .

You can reset bios and see if it is ok after that.

If you don't want to reset bios then it is harder to find the problem.

If you did bclk overclock then your iGPU is overclocked also, and you need to change the iGPU multiplier in bios, by setting a lower frequency, and increase VCCIO voltage.

If you did a CPU multiplier overclock then you need higher voltage on VCCIO, because iGPU use it also.

[CKowa1]

Hey Ronin, thanks for such fast response. I forgot to mention, that I was trying to go back from OC to stock, and that did not help at all. As for my OC, I'm using Game Bost feature on my motherboard which also increases ram and cpu voltage. Do I also have to look after iGPU multiplier and voltage? Also maybe updating bios will help?

I'm waiting to get gtx1070, so few more days of pain, and I hope everything will get back to normal with new third party gpu, as it was normal with gtx970

[CCris2]

Yes, you should update to the last bios available for that mainboard.

[DHahn]

Hello everyone,

I'd would also like to join the club in this thread. First of all my system:

i7 6700K not overclocked

ASUS Z170-A newest BIOS version

32GB 2133MHz of supported Corsair memory

2x R9 390 with newest drivers

I should also mention that I had to replace the CPU once already, as it was not able to boot any operating system anymore after a few weeks. I got the Machine_Check_Exception bluescreen. Only deactivating 3 Cores and Hyper-threading would boot back to Windows. So I replaced the CPU with a new one which fixed the problem. (old one was from week 34 2015 or something, the new one is from week 2 2016)

Since then I'm having the same issues as mentioned above. My system is stable with heavy workloads. For example: Prime was running all night even with the i7 OCed only via the multiplier to 4.5GHz. Anyway rather small workloads like playing World of Tanks, Browsing or even at idle my computer will freeze completely. I also noticed that the computer will freeze more frequent after a gaming session and just simple browsing afterwards for example. So my first guess would be a problem with Speedstep/C-states or something like that.

I nearly tried everything with different drivers, different BIOS versions and settings, etc. The last two things I can do is turn off C-states and look what happens and on the other side use the i5 6600K my girlfriend uses in her setup. Both options would indicate to switch the i7 once again I guess if it works then.

Is there anything else I didn't think of? I have to admit I didn't read the complete thread here. I've been reading all over the internet since I couldn't find any fix for it so far. Maybe someone here can give me a hint because I missed the solution? Or is the last status: "Intel is checking the problem, no solution except another change of CPU so far"?

I'm looking forward for every answer. Thank you in advance for every help.

[MBell11]

Hi,

When you say your cpu is not overclocked, how is the vcore defined on ? Auto ? Manual ? Offset ? Adaptive ?

My cpu was unstable on idle with adaptive voltage because I had forgotten to set a positive offset to prevent the vcore to drop too low on idle and light loads. I added a 0.03v offset and now it's stable (the vcore drops to 0.816ish instead of 0.784ish on idle).

[DHahn]

Hi,

thanks for your answer. Everything was set to auto. HWMonitor tells me Min: 0.696V, Max: 1.408V, but for some reason I never see this value altough CPU-Z and HWMonitor tell me the CPU is running at 800MHz. The Core Voltage seems to stay at 1.392V. Which program would you suggest to read out the VCore?

Edit: Alright, I found a option in BIOS which prevented the VCore to change. The system was always running with this 1.392V according to CPU-Z. After I changed that option the VCore dropped to 0.784V just like yours. Now I just set an offset of +0.1 and it doesn't drop below 0.880V. I will check this now. If it works now I'm fine altough that would be rather strange if the VCore never dropped before. I'm quite sure I had this setting for quite a while but just didn't notice it. By the way, don't worry about temperatures under load. They are no problem as everything is custom watercooled, with lots of case fans still moving air in the case.

I'll be back with a result as soon as it freezes again, if not I will also try to let you guys know.

[MBell11]

I use CPU-Z to read the vcore. Be careful with offset, because it also increases the voltage under load

Edit : you said not to worry about temperatures, but high voltages, even if you have good cooling are not harmless...

[DHahn]

Like I said, temperature is no problem. Everything watercooled. So the offest didn't help at all. World of Tanks crashed after like 2 minutes.

Right now I turned C-states off and also the voltage regulation so the processor switches between different clocks but has always a stable voltage. So far it looks stable...

I already had nearly the same settings applied, the only difference right now is the turned of C-states, so that seems to be the problem. But I'll keep playing for a little longer to verify my assumption.

Edit: Yeah, of course they are not harmless, but the VCore was still lower than right know with the automatic fixed VCore of 1.392V. At least that's what CPU-Z reads.

Edit2: Ok, so I take everything back... The computer just froze again with more or less no load. I was just chatting with someone via Steam and the Browser was opened... And right after rebooting it froze a second time... Any other ideas?

[DHahn]

Hello everybody,

good news in my case: The system was now stable the last three days. Here is what helped with my problem:

I set the RAM voltage to 1.35V which is the value for the overclocked XMP profile. This didn't help that much. But after reducing the clock speed of the RAM from the 2133MHz set by the XMP profile to 2000MHz the system didn't freeze a single time anymore. The first XMP profile are the values the RAM should manage without a problem, so I guess I'll return it...

My 6700K is now running with everything set to auto, C-states enabled, Speedster enabled, etc.

If anything changes I'll post again. I'll also check back when I have some new RAM eventually.

I hope this helps at least some of you.

[CCris2]

Hi eyeless45 ,

First of all set vcore to Adaptive.

Changing RAM is not a solution. I don't know what is the exact RAM specification, because I don't know the model you have.

For more memory stability you should increase VCCSA and VCCIO. Try for example VCCSA 1.2V and VCCIO 1.15V and check stability after that.

In this thread, if you are patient to read it, sahafiec changed 3 CPUs and 3 types of memory on ASUS mainboard and in the end he modified voltages and no more system crashes :

It seems that mainboard is setting wrong voltages.

[DHahn]

Hi Ronin,

thank you very much for your advice. I have a 32GB Corsair 2133Mhz CL13 Kit which is also on the QVL list of the mainboard. Setting the XMP profile 1 (2133MHz, CL13, 1.2V, etc.) and setting VCore to Adaptive already helped quite a lot, but the system wasn't 100% stable. So after that I set VCCSA and VCCIO both to 1.15V. Since then I had no crashs.

[CCris2]

You can lower those voltages till you see that your system remains stable. This was supposed to be asus bios job, I mean the bios to set automatically those voltages, if the memory is in QVL ...

[CKowa1]

Hello again. It's been awhile since my last post @ 03.06. After Ronin's advice i went back to stock cpu speed once again. Since then i had only 3 freezes. First two were like 3 days ago and both in like 10 minutes period. Last one i had today. I was working on stock cpu speed and xmp @2400mhz with stock voltage. After todays freez i changed to xmp 2666mhz and voltage of ram automaticly was incresed to 1,3 from 1,2. Before those three freezes I was able to play Far Cry 1 on my onboard gpu like 10 hours in last few days and nothing happend. Freezes always come when browsing websides, changing winamp song, moving files from dashcam to hdd...

[CCris2]

Hi Makabra ,

Set your memory to auto and update your bios from here: https://www.msi.com/Motherboard/support/Z170A-GAMING-PRO-CARBON.html# down-bios Support For Z170A GAMING PRO CARBON | MSI Global | Motherboard - The world leader in motherboard design

[CKowa1]

Today I finaly did bios update after two more freezes when I was watching YT. So i can say that xmp @2666mhz did not help Anyway, I cannot belive that all my problems started after selling gtx970 and jumping on to integrated graphic card.