After a Windows 10 Update Today Overclocking is lost. WTF Microsoft and Intel???

[RejZoR]

I had to downgrade my MSI X99A Gaming 7 bios because it was pissing its own pants non stop. After downgrade, problems were gone. No more random bsod-less crashes and random startup bsods.

 

[cadaveca]

Sounds like something pulling the cmos battery/shorting clr cmos pins might solve... but it also seems like a weird design flaw (choice)?

Unfortunately, no, CMOS clear does not help. Something gets corrupted (microcode, or MEI perhaps?), and the only way to fix it is to reflash the BIOS. Tested with my own review sample boards, with retail boards in my possession, other users out there on the internet into memory clocking... and several users here on TPU. That's why, as soon as I hear OC not working any more, I suggest reflash BIOS, and a fresh OS install.

Think about that for a second though, and what I am saying... how could an OC affect BIOS? What part of reading a BIOS to boot corrupts it? Should this not be impossible?

Give me nearly any ASUS motherboard, and I can force this to happen, and you can search yourself and find users that have a seemingly perfectly working board, but things like MEI version listed in BIOS disappears... I've been almost a singular voice about CPU OC degradation, and many people have discounted the idea, but this is something I have spent hundreds if not thousands of hours testing and investigating.

I'm not the only person aware of this specific issues with ASUS boards though.... and the idea that Windows cannot force a BIOS flash in incorrect.

But lets start with someone else who knows about this too... such as this thread from 5 years ago that describes a recovery method if gone too far:

https://www.overclockers.com/forums...o-fix-your-BIOS-if-there-is-corrupted-ME-area

There is much more to this topic that can be discussed, but I'm not sure this thread is that place.

 

[hat]

Unfortunately, no, CMOS clear does not help. Something gets corrupted (microcode, or MEI perhaps?), and the only way to fix it is to reflash the BIOS. Tested with my own review sample boards, with retail boards in my possession, other users out there on the internet into memory clocking... and several users here on TPU. That's why, as soon as I hear OC not working any more, I suggest reflash BIOS, and a fresh OS install.

Think about that for a second though, and what I am saying... how could an OC affect BIOS? What part of reading a BIOS to boot corrupts it? Should this not be impossible?

Give me nearly any ASUS motherboard, and I can force this to happen, and you can search yourself and find users that have a seemingly perfectly working board, but things like MEI version listed in BIOS disappears... I've been almost a singular voice about CPU OC degradation, and many people have discounted the idea, but this is something I have spent hundreds if not thousands of hours testing and investigating.

I'm not the only person aware of this specific issues with ASUS boards though.... and the idea that Windows cannot force a BIOS flash in incorrect.

But lets start with someone else who knows about this too... such as this thread from 5 years ago that describes a recovery method if gone too far:

https://www.overclockers.com/forums...o-fix-your-BIOS-if-there-is-corrupted-ME-area

There is much more to this topic that can be discussed, but I'm not sure this thread is that place.

Something to keep in mind in case I ever happen across such weirdness.

 

[rkt]

This thread is only 6 pages long at this point and there are at least three solutions already stated here.

Have you read it?

After the update KB4100347 was uninstalled - the system was recovered (also overclocking ability), so I paused windows updates for 35 days, but what next? Try to hide this update?

 

[Candor]

After the update KB4100347 was uninstalled - the system was recovered (also overclocking ability), so I paused windows updates for 35 days, but what next? Try to hide this update?

Use this tool to hide the update (will prevent it from being re-downloaded).

http://download.microsoft.com/download/f/2/2/f22d5fdb-59cd-4275-8c95-1be17bf70b21/wushowhide.diagcab

Instructions...

https://www.thewindowsclub.com/block-unwanted-windows-updates-in-windows-10

 

[rkt]

Use this tool to hide the update (will prevent it from being re-downloaded).

http://download.microsoft.com/download/f/2/2/f22d5fdb-59cd-4275-8c95-1be17bf70b21/wushowhide.diagcab

Instructions...

https://www.thewindowsclub.com/block-unwanted-windows-updates-in-windows-10

Thank you! Done. Till next malicious update

 

[Candor]

You're welcome. Begone crappy update.

 

[R0H1T]

That's like delaying the inevitable, will any of you skip the entire RS5 or RS6 updates next?

 

[RejZoR]

Use Feedback Hub in Win10 to report this in detail. Then we can upvote the issue to make it noticeable. That's the best way. Also mention KB number in the title as a reason of problem.

 

[freeagent]

Unfortunately, no, CMOS clear does not help. Something gets corrupted (microcode, or MEI perhaps?), and the only way to fix it is to reflash the BIOS. Tested with my own review sample boards, with retail boards in my possession, other users out there on the internet into memory clocking... and several users here on TPU. That's why, as soon as I hear OC not working any more, I suggest reflash BIOS, and a fresh OS install.

Think about that for a second though, and what I am saying... how could an OC affect BIOS? What part of reading a BIOS to boot corrupts it? Should this not be impossible?

Give me nearly any ASUS motherboard, and I can force this to happen, and you can search yourself and find users that have a seemingly perfectly working board, but things like MEI version listed in BIOS disappears... I've been almost a singular voice about CPU OC degradation, and many people have discounted the idea, but this is something I have spent hundreds if not thousands of hours testing and investigating.

I'm not the only person aware of this specific issues with ASUS boards though.... and the idea that Windows cannot force a BIOS flash in incorrect.

But lets start with someone else who knows about this too... such as this thread from 5 years ago that describes a recovery method if gone too far:

https://www.overclockers.com/forums...o-fix-your-BIOS-if-there-is-corrupted-ME-area

There is much more to this topic that can be discussed, but I'm not sure this thread is that place.

Clearing the cmos had no effect, even pulling the battery only partially clears the cmos, This is the first board that I've had that does that.. For me the timing was weird, the update, pushing my mems, and then no stable oc. My R3F would never think of shenanigans like that lol. Well maybe it would, it can be a bitch to.

 

[lexluthermiester]

This update seems to be the gift that keep on giving.

 

[DRDNA]

7700HQ and GTX1070 with 16 GB ram all current patches to my unnamed and mystery built laptop ( this make of laptop is not aloud to be mentioned on the TPU website ) are working 100% with no slow downs.Benched before and after installs and all is with in margin of error.

BEFORE PATCH Benches

AFTER PATCH

BEFORE PATCH

AFTER PATCH

Before PATCH

AFTER PATCH

Also worth a mention is XTU was NOT affected and the undervolting is still working fine with no issues and the patches installed with out any issues either and all restarts for installation completion were also 100% successful.

 

[bug]

Well f*uck me. Forced uninstalled that update in safe mode and now system is stuck in BSOD loop and can’t even get into recovery mode!

Thanks a bunch Microsoft for mess up my work computer with your “Intel validated” microcode update.

They never claimed they have validated the downgrade

 

[cucker tarlson]

all those updates got me worried,to the point that I decided to dig out an old thread and test my 5775c in the heaviest gaming scenario I can think of. Happy to report no perf was lost.
https://www.techpowerup.com/forums/...ct-on-gaming-performance.236514/#post-3716262

now I get the same 118 fps

https://imgur.com/a/SIy6nJ0

 

[rkt]

That's like delaying the inevitable, will any of you skip the entire RS5 or RS6 updates next?

Maybe it’s time to switch to AMD?

 

[cucker tarlson]

Maybe it’s time to switch to AMD?

In HEDT it's a no brainer.

 

[rkt]

In HEDT it's a no brainer.

But there is no clear way how to avoid such serious system damage?

 

[cadaveca]

But there is no clear way how to avoid such serious system damage?

X299 is not affected by this problem that I can tell, but BIOS development for that platform is still active at most brands.

Maybe it’s time to switch to AMD?

That's really not an easy question to answer. I have been running HEDT both platforms side-by-side since last year and there are benefits to each.

 

[rkt]

X299 is not affected by this problem that I can tell, but BIOS development for that platform is still active at most brands.



That's really not an easy question to answer. I have been running HEDT both platforms side-by-side since last year and there are benefits to each.

Maybe to wait for refresh (due to overheating)?

 

[RejZoR]

Lets all remember this gem I've encountered some time ago:
https://www.techpowerup.com/forums/...or-deliberate-performance-degradation.246065/

Basically same scenario and how everyone got all sorts of ideas and theories that it's "because that A9 is just crap" and how somehow didn't know something even though I had numbers plain visible even in something as basic as CPU-Z benchmark. History repeats and will continue to repeat apparently.

 

[adamwzl]

I have to agree with the OP. Using a ASUS Rampage V Edition 10 and 6950x with update KB4100347 installed causes Overclock to disappear. Only way I was able to get it back with the update installed was to use Intel XTU and set my core clock there. All voltages and DRAM settings stayed from BIOS only difference was the core clock values. But with doing that it introduced stability issues. Overwatch and CSGO keep crashing consistently causing suspensions and rank loss on competitive. Once I removed the update all those problems went away. I'm still using the BETA 1902 bios from ASUS and my Overclock is staying just fine @ 4.3.

Hopefully ASUS reverses it's decision and releases a non BETA version of BIOS to support the latest Intel microcode update. X99 is not that OLD yet, all processors are still capable workhorses.

 

[mkdr]

@DRDNA ... you already had a MC and Meltdown+Spectre protections enabled, so of course, there is no effect on you on this update...

 

[remixedcat]

even disabling service does nothing it reinitializes itself in a day

 

[rtwjunkie]

This is why the Windows Update Management Tool is such a good little program. It allows you to block updates just like previous Windows versions.

 

[RejZoR]

This is why the Windows Update Management Tool is such a good little program. It allows you to block updates just like previous Windows versions.

When it works. There are also situations where something will get installed regardless. It's especially bad with drivers that seem to be pulled from some parallel channel...