Intel Skylake bug seizes up PCs running complex workloads

Well, isn't this fun...

Intel Skylake bug seizes up PCs running complex workloads
http://hexus.net/tech/news/cpu/8963...loads/?utm_source=dlvr.it&utm_medium=facebook

" BIOS updates on the way to fix problem says Intel
Four days ago Intel reported that its engineering department had identified the issue which "only occurs under certain complex workload conditions... [when] the processor may hang or cause unpredictable system behaviour". It has released a fix for the issue to hardware partners which will be distributed via BIOS updates for Skylake compatible motherboards. Now users will just have to wait for their motherboard vendors to publish BIOS updates with the Intel fix incorporated."

Incoming Intel Errata patch, BIOS update imminent

Update: According to post #323:
http://www.mersenneforum.org/showpost.php?p=421898&postcount=323

"The microcode update 0x6A (or higher) should solve the problem. We will see when the new Bios is released."


Related News:
There seems to be another Skylake instability issue that's been going on for a while, and was recently reported to Intel, here are the links.

Most recent post is a summary:

New Skylake Build Randomly Freezing/Crashing - Page 8
http://www.tomshardware.co.uk/forum...ndomly-freezing-crashing/page-8.html#17336235

Skylake hard locks seeminly when IDLE / CStates ON
https://communities.intel.com/thread/97358

Should we start a new top level thread, or track it here?

Anyone seeing similar symptoms with their Skylake laptop / desktops?

 

Yeah and for people with a Clevo based laptop, good luck having Clevo release any fix for anything! Heh!

 

Found in November, discussed with Intel mid December. Not so many people noticing it then again not to much SW using FMA.

 

That is really ugly - many vendors will obviously skip the updates on some of their models.

 

GT72S 6QE - microcode update folded in to E1782IMS.112
http://www.msi.com/product/notebook/support/GT72S-6QE-DOMINATOR-PRO-G.html#down-bios

"Description Fix ODD AutoRun abnormal issue.
Note To complete the Processor Microcode update, please follow the SOP below to perform BIOS update twice with UEFI BIOS Utility."

Originally posted by Kevin@GenTechPC
http://forum.notebookreview.com/thr...n-owners-lounge.769092/page-588#post-10176377

 

Someone must have misunderstood the MSI update...It is from mid December, using Intel code from September and doesn't fix a thing related to the "bug"...sorry.

 

I thought the problem was found in October, solved and discussed with Vendors in November, so wouldn't December be a good first potential time for an update?

The publish date is 2016-01-05, and I assume MSI was testing through the holiday to make sure it was solid, and maybe didn't want to push the update out until the problem was announced by Intel recently.

The timing fits, why don't you think it is the Skylake microcode update?

@Kevin@GenTechPC did you confirm this was the Skylake bugfix?

 

@Mr. Fox (reg.: v1.1)
We are using the patch in the Clevo Mods...
I took the MSI update apart to check for you guys and the patch is not present...sorry.

 

Which Clevo BIOS contains the patch? 1.05.03?

 

Thanks for checking Prema, I updated @Kevin@GenTechPC :
http://forum.notebookreview.com/thr...n-owners-lounge.769092/page-588#post-10176471

 

As of P870DM-G Clevo stock BIOS 1.05.04 it doesn't contain the patch...
Only the v1.1 Prema Mod version does...with brother Fox in beta testing since last week.

 

More background information / links:

Simple instructions for freezing a Skylake Processor.
https://communities.intel.com/mobile/mobile-access.jspa#jive-content?content=/api/core/v3/contents/524553

768k Skylake Problem/Bug
http://mersenneforum.org/showthread.php?t=20714

According to post #323:
http://www.mersenneforum.org/showpost.php?p=421898&postcount=323

"The microcode update 0x6A (or higher) should solve the problem. We will see when the new Bios is released."

According to post #326
http://www.mersenneforum.org/showpost.php?p=421916&postcount=326

"Those who found the bug, and then those who were able to reproduce it, found that hyperthreading had to be enabled. Also, it only manifested when Prime95 / mprime was running a 768K FFT using the AVX code path.
However, Intel obviously have a much deeper understanding of the issue, so it's possible that other software might encounter the issue without hyperthreading enabled.

 

I ran the test as described here for 1 min, all stock clocks/ voltage and my system didn't crash. I only ran it for 2 mins though

/using CLEVO BIOS 1.05.02 / PREMA VBIOS for the GPU(s)

 

Apparently it can take many hours to manifest:
"- Let the program run for some time and at some point, minutes or hours, the system will freeze."

If it ran for 48-72 hours I would think it was cured... at least 98% sure

 

Will there be a Prema bios for the Clevo N240JU that fixes the skylake bug?

 

Guess I won't be using Prime95 to benchmark my notebook anytime soon....

 

Set set it up exactly as explained in the Simple instructions for freezing a Skylake Processor thread and I ran Brother @Prema latest beta BIOS v.1.1 mod with Prime95 on the Sky X9 from 4:30 PM to 7:45 PM without so much as a hiccup on the 6700K. So, if you are using a Skylake machine with Premamod support you should not have anything to lose any sleep over. Not sure if it was actually important in the first place since it never affected me before, but if it was, it's not now. Good job, Prema.

I never have viewed it as a benchmark tool and never really found it to be much good for anything except bringing happiness to people that enjoy torturing CPUs with no purpose anyway. I love benchming but just making my stuff get hot for the hell of it... I don't get that. Never understood that with Furmark either.

I don't think I have left a computer turned on for that long in years. When I want to use a computer I turn it on. When I am done I turn it off. When I want to use it again, I turn it back on again... just like my washing machine, clothes drier, dish washer, car, weed eater, and my lawn mower... *deep breath* ...soldering iron, electric drill, table saw, air compressor, and so on and so forth. But, I don't run a server or do any web hosting so I don't need to wear 'em out when not using them.

 

0x6A is dated 14-Dec-2015 so doesn't make sense that it would include the fix as that is the date the problem was first informed to Intel who didn't respond until the 16th for more information and then only acknowledged providing a fix on 6-Jan-2016.

EDIT: Seems Intel had already fixed the bug with 0x6A before the boys at mersenne brought it up at the Intel Community forum. :/

 

There seems to be another Skylake instability issue that's been going on for a while, and was recently reported to Intel, here are the links.

Most recent post is a summary:

New Skylake Build Randomly Freezing/Crashing - Page 8
http://www.tomshardware.co.uk/forum...ndomly-freezing-crashing/page-8.html#17336235

Skylake hard locks seeminly when IDLE / CStates ON
https://communities.intel.com/thread/97358

Should we start a new top level thread, or track it here?

Anyone seeing similar symptoms with their Skylake laptop / desktops?

****Update: started new thread****

Skylake hard locks when IDLE / CStates ON
http://forum.notebookreview.com/threads/skylake-hard-locks-when-idle-cstates-on.786966/

 

Different issue, I'd make a new thread.

Seems Intel should have held back Skylake... Wonder what else will be discovered broken.

 

Hi yes skylake cpus been manufactured with error ..
Intel claims that it can be solved by updating BIOS they seeded for every laptop or desktop manufs or MB manufs.
But how can BIOS update solve CPU error problem, have no idea..
Intel hiding something.. or they admitted ? found link..
http://www.theregister.co.uk/2016/01/11/math_bug_splatters_skylake_intel_working_on_fix/

 

Yup, if you want to avoid the 2 Skylake bugs now, before new microcode based BIOS updates arrive:

1. C-States => OFF - stops Idle hangs.

2. Hyperthreading => OFF - stops MATH computational hangs.

Skylake, bringing you more, with less

 

If you want to avoid all bugs

1. Power => OFF - stops all bugs, malware, virus.

BTW microcode ver 0x74 is out already and can perhaps be temporarily applied through OS or UEFI before a new BIOS is released.

 

Hello,

I have a MSI GT72 6QD Skylake 6700, BIOS E1782IMS.112 / EC 1782EMS1.109 and since i have it i have some random weird errors :

Random limited connection in ethernet rj45
Error corrupted for random files when installing big program (never same files after a reboot)
Ftp random drop error gnu-24 decryption failed

but when i uninstall nvidia driver with ddu everything work, all these problems disappear (i tried different versions of nvidia drivers ofc)

I have win 10 with last updates,
It happen also with Windows 8.1 or even with 7 so i guess it's not driver (and i tried driver from msi and from editors last updates)

could this be bios related?

 

And how would you apply this through the OS? Just wondering as it makes no sense to me how it's possible.. No offense or anything meant, just genuine curiousity on my part..

Sent from my LG-H811 using Tapatalk

 

Click Start, Power, Shutdown.



Oh, you mean the other part?

 

I meant the other part.. of course I know how to restart tiller, you cheeky bugger!

 

Can't tell too easily with someone from London...

 

It's just a patch. If you go into your BIOS and check your microcode version and then check it in Windows using HWiNFO is it the same? It's not uncommon for Windows to apply a higher patch version.

Many ways to update but they all basically use the same method. If someone with a Skylake would like to update to 0x74 without touching the BIOS let me know and I'll try to post something that might be usable.

 

I have actually c states on and all is ok ..
hyper thrd-ing off no need as not doing anything heavy then it saves power and want my laptop as quiet as poss. and cool as poss. when browsing net or emailing or doing office work..

I had something changed previously in BIOS and post took longer so black screen for few more seconds then msi appeared.. so I did defaults and couple minor changes and all OK now

 

How does disabling HT save power?


Sent from my iPhone using Tapatalk

 

CPU threads are not used all in same time.. less power and less heat..
when you play gamez turn HT off
when you editing videos or compression task .. or virtual machines turn it on ..
The label ''hyper'' says all cores goes hypo when on

 

People using Windows who want microcode ver 0x74 update for their system now can find a solution here.

Ver 0x6A was supposed to take care of the bug and 0x74 a few things more I guess.

 

See:
http://www.tomshardware.com/news/asrock-releases-bios-updates-to-fix-intel-skylake-bug,31063.html


Fixes are coming since the 18th it seems.

 

Looks like Intel is focused on more important issues.

http://www.pcworld.com/article/3031...ocking-party-by-closing-skylake-loophole.html


Sent from Mr. Spock through Tapatalk

 

Wild, I thought a firmware update was coming, but we still haven't seen it for the MSI GT80S or GT72S. I guess this is why...they are still working on rolling out a fix.

How to test your PC for the Skylake bug
Intel confirms there's a bug that could crash Skylake CPUs, under specific and somewhat uncommon circumstances.
Gordon Mah Ung | Executive Editor, PCWorld Jan 11, 2016 1:19 PM
http://www.pcworld.com/article/3021023/hardware/how-to-test-your-pc-for-the-skylake-bug.html

"The bug initially was reported only with the Core i7-6700K desktop CPU with Hyper-Threading enabled, but Intel’s confirmation seems to indicates it impacts more CPUs in the lineup."

" The bug has apparently been stewing for weeks on forums at hardwareluxx.de and then Mersenne.org which created the software, Prime95, that is used to induce the bug. Prime95 is used to find prime numbers and is also very popular with performance and the overclocking community as a stress and performance test."

"Besides the community post, Intel also confirmed to PCWorld the existence of the bug but placed an emphasis on the word “ might” because there’s no guarantee you’ll hit the bug.

“Under some complex workload conditions, like those encountered when running applications such as Prime95, the processor may hang or cause unpredictable system behavior. Intel has released a fix that resolves the issue and we are working with external business partners to deploy this fix through BIOS updates,” an Intel official told PCWorld."

The article then goes into some detail on how to set up and test for the bug.