Processors
Intel® Processors, Tools, and Utilities
14400 Discussions

Solution asked for failure C3/C6/7 states - 6700K CPU

ALips
Beginner
3,555 Views

My apology for opening a new thread about this issue, but I have several questions on this matter, hopefully from someone who is from Intel who can give a answer.

Short case:

C-states are not functional as intended on a Skylake CPU in combination with Windows Pro 10/Linux, so far I can tell from reading different topics on the internet all cpu's are involved, all Z170 chipsets are involveld, all motherboard manufactors are involved. When you enable state C3/C6/C7 in the bios with differnt microcode from Intel [26-01-2016] the result is a random freeze [no BSOD] on the OS.

https://bugzilla.kernel.org/show_bug.cgi?id=109081 https://bugzilla.kernel.org/show_bug.cgi?id=109081

http://www.tomshardware.co.uk/forum/id-2830772/skylake-build-randomly-freezing-crashing.html http://www.tomshardware.co.uk/forum/id-2830772/skylake-build-randomly-freezing-crashing.html

https://rog.asus.com/forum/showthread.php?77522-Maximus-VIII-Hero-Freezeing-(ACPI-error-) https://rog.asus.com/forum/showthread.php?77522-Maximus-VIII-Hero-Freezeing-(ACPI-error-)

So I have a few simple questions for Intel:

  1. Is Intel aware of this situation?
  2. Is it a hardware problem, aka - is it neccesary to replace the cpu or is it something which can be solved through a updated microcode?
  3. Is Intel working on a solution to fix this when this can be solved through a microcode update?
  4. Is it possible to give some more details about this situation?

I'm asking this because I bought 9 days ago a compleet new system based on a Skylake system and all what I'm doing now is testing/avoiding a freeze. The supplier of the cpu is aware of situation and wants to know if I want to replace the cpu, ASRock is also helpfull with toughts or views to fix this situation, but with no luck so far.

I really hope that someone from Intel give me and other people some news what to espect in this situation.

Thanks for your time reading.

Het bericht is bewerkt door: Arno Lips Added C3 state, just got freeze with only C3 enabled....

0 Kudos
8 Replies
MV12
Novice
2,015 Views

I would like to second these 4 questions

  1. Is Intel aware of this situation?
  2. Is it a hardware problem, aka - is it neccesary to replace the cpu or is it something which can be solved through a updated microcode?
  3. Is Intel working on a solution to fix this when this can be solved through a microcode update?
  4. Is it possible to give some more details about this situation?

I have Skylake 6600 and before I switched the C-stats to [Disabled], I was experiencing complete freezes (both Win 10 and Linux) between 1 and 3 times a day - sometimes when the PC was completely idle, and sometimes during some ordinary tasks, like web browsing, but curiously, never when running games.

A lot of people from that Tom's Hardware thread are experiencing the same, few have tried to install the new BIOS from their motherboard manufacturers which includes the fix for the Prime95 problem, but the freezing issue still persists, so it is a separate problem.

Amy_C_Intel
Employee
2,014 Views

Hello all,

I would like you to refer to the following post where this issue has been addressed.

Regards,

0 Kudos
ALips
Beginner
2,014 Views

Hi Amy, thanks for your response.

Are you reffering to the Mersenne Prime Search error? In that case I tought the solution was allready give in microcode 6A/64.... even with microcode 74 the system freezes.

As stated in other posts the freeze is not caused by heavy workload, but when the system is almost idle..... like playing music, browsing, etc... cpu usage 1%

So can you be more specific why you're are reffering to that post?

And so the orignal questions still apply:

  1. Is Intel aware of this situation?
  2. Is it a hardware problem, aka - is it neccesary to replace the cpu or is it something which can be solved through a updated microcode?
  3. Is Intel working on a solution to fix this when this can be solved through a microcode update?
  4. Is it possible to give some more details about this situation?

Shoudn't be that hard for a compagny like Intel to give some answers on some simpel questions...

Regards,

0 Kudos
Amy_C_Intel
Employee
2,014 Views

The answer to your questions are answered on the post I referred, in which the answers are provided by our engineering personnel.

 

  1. Yes, Intel is aware.
  2. This issue only occurs under certain complex workload conditions, like those that may be encountered when running applications like Prime95.
  3. Intel has identified and released a fix and is working with external business partners to get the fix deployed through BIOS.
  4. No further details have been provided yet, all the information available in regard to this matter are stated on this post .

Regards,

0 Kudos
ALips
Beginner
2,014 Views

Hi Amy, Thanks for your response.

So first let me start quoting myself and pay attention to the underscores.

Arno Lips schreef:

Are you reffering to the Mersenne Prime Search error? In that case I tought the solution was allready give in microcode 6A/64.... even with microcode 74 the system freezes.

As stated in other posts the freeze is not caused by heavy workload, but when the system is almost idle..... like playing music, browsing, etc... cpu usage 1%

So after I read your reply the system got frozen again...doing nothing! Only Chrome open and your answer to this topic. Structed with the irony of the situation, I had a good laugh!

Let's makes this simple and we follow the bullets:

  • Yes, Intel is aware. - Read the quote again! and then I refraise the question for you.... Is Intel aware that there is a other bug/error beside the Mersenne Prime Search error? Like that when C-states are enabled and the cpu load is 1% and idle - the system has a lockup.

  • This issue only occurs under certain complex workload conditions, like those that may be encountered when running applications like Prime95. - Like I stated before, I'm allready running the fix x074 updated in the bios from ASRock. But hé... sometimes a picture can see more then a thousand words.

The source with instructions how to test - http://www.pcworld.com/article/3021023/hardware/how-to-test-your-pc-for-the-skylake-bug.html How to test your PC for the Skylake bug | PCWorld

And while we at it, a screenshot with a memorytest wich is running triple the time to establihs that it's stable.

 

  • Intel has identified and released a fix and is working with external business partners to get the fix deployed through BIOS. - Thanks for the hard work on the Mersenne Prime Search error, I did recieve the fix. Can we now focus on 'Help I'm going to the toilet and have to save all my work because I not shure if my new pc is working when I'm coming back' - 'bug/error'

  • No further details have been provided yet, all the information available in regard to this matter are stated on this post/thread/96157 Simple instructions for freezing a Skylake Processor. - Well Amy, no offence, but I have a simpler bug. I only have to open a browser or playing some music to let the system crash with C-states enabled. I really don't have to startup Prime95 and adjust a config.txt and wait for two hours, It can happen every moment!

So enough with the irony and sarcasme and can you put me in contact with somebody from Intel who can give some directions how to proceed to solve this issue. I really would apreciat that!

Thanks for your time reading.

Regards,

0 Kudos
GGlyn
Novice
2,014 Views

Please also refer to the original thread that I posted.

THIS CState/IDLE error has nothing to do with Prime95. If there is any confusion in that regard, please stop, read and then push the right people. If you keep believing it is Prime95 related, there will never be anyone with any real technical expertise hearing about it or resolving it.

Again. I realise the temptation to assume all problems are that one problem (and I have been involved in support situations long enough to tell you I understand why you think all Skylake issues are Prime95) but this issue here is NOT the same issue. The only people to really notice the Prime95 issue of which you feel is so widely affecting people, is tiny. Hell, the average PC owner could use a Skylake, passed down over generations until his great great great granddaughter was using it one day and wondered why it crashed!!! Its that unlikely... Unless of course Intel is not being 100% open about the "very specific workload" discussion

Prime95 has nothing to do with CState mechanics. CStates disabled would NOT resolve that bug.. but CStates disabled resolves THIS bug. Prime95 problems are under load by definition.. (to run Prime95 creates load). This is exactly the opposite.. WITHOUT load it happens.. with LOAD it does not... They are different in symptom, response, diagnosis.. everything...

0 Kudos
KTran10
New Contributor I
2,014 Views

This time Intel need to acknowledge there is a C-State bug of Skylake and fix it. I speculate Intel has quality control problem lately. They have not got the yield they need, so someone cutting corner.

0 Kudos
MV12
Novice
2,015 Views

Dear Amy,

Thank you for replying, however I have to say that I am fairly disappointed in the manner of your replies. It has been stated with pristine clarity in the Original Post, and in my reply to it, that the issue reported is nothing to do with Prime95 bug acknowledged by Intel and described in the thread you have quoted, , but that it is to do with C-states, as there is a fairly large body of evidence from the people who have researched this matter extensively that the issue is not to do with complex overloads and mathematical computations. In fact, the chronology of this thread is:

1. The original poster goes to great lengths to provide as many details as possible to describe the problem with C-states and Skylake CPUs freezing

2. I explicitly say that I was experiencing this problem on a daily basis until I set C-states to disabled, which seems to have permanently solved the problem

3. You reply quotes the Prime95 bug thread, without even mentioning C-states or anything in the original problem description.

4. Another poster states that the issue is not with Prime95 big, but with C-states, and that freezing happens during completely idle times as well

5. You reply quotes the same Prime 95 bug thread, without even the slightiest of acknowledgements that you have even read the description at all, as again, you fail to mention anything to do with C-states, and keep repeating that the problem is resolved, and that the freezing occurs under "certain complex workload conditions".

I'm sorry, but no, it does not.

The problem that we are discussing in this thread, which is nothing to do with Prime95 bug Intel already acknowledged and resolved, is to do with Skylake CPUs experiencing freezes under any conditions, on a wide range of motherboards from different manufacturers. For a vast majority of people from the links provided in the above posts, changing the configuration of C-states, or in my case, disabling the C-states altogether has vastly improved the stability, or fixed the problem altogether.

However, disabling something is just a workaround, and we would all like to know at least:

- Is this problem, which is nothing to do with Prime95 bug Intel already acknowledged and resolved, anywhere on Intel's list of known issues?

- Is it a hardware problem, aka - is it neccesary to replace the cpu or is it something which can be solved through a updated microcode?

- Is Intel working on a solution to fix this when this can be solved through a microcode update?

- Is it possible to give some more details about this situation?

We (speaking for myself, at least) do not expect answers within the same 30 minutes, we do not expect possible BIOS updates and microcode fixes in a matter of days or weeks even, but we would dearly appreciate resourceful information from Intel about a problem which very obviously exists.

Thanks

0 Kudos
Reply