Thursday, October 21st 2021

Windows 11 Performance Issues on Ryzen Fixed by Updates from Microsoft and AMD

by
btarunr
 Discuss (85 Comments)
Microsoft and AMD on Thursday released software updates that fix the two performance issues affecting AMD Ryzen processors with Windows 11. The two issues were abnormally high L3 cache latency, and a broken "Preferred Cores" system. The companies had assessed that the issues impact performance of Ryzen processors on Windows 11 by as much as 15%.

The two issues are fixed in separate methods. The L3 cache latency bug is improved through a Windows Update patch, which has been released now as an Update Preview (an Update Preview is not a "beta," but a software update released ahead of its designated "patch Tuesday"). The Update Preview is chronicled under KB5006746, and Windows 11 systems updated with this, get their OS build version set as "build 22000.282." The next update restores the Preferred Cores mechanism that leverages UEFI-CPPC2. This update comes in the form of an AMD Chipset Software update. You'll need to download and install both of the following:

DOWNLOAD: Windows 11 October 21, 2021 Update Preview KB5006746 | AMD Chipset Driver Software 3.10.08.506

Related News

Add your own comment

85 Comments on Windows 11 Performance Issues on Ryzen Fixed by Updates from Microsoft and AMD

#51
KarymidoN
TheoneandonlyMrKIs it still a free upgrade from 10pro?!
yep
Posted on Reply
#52
LiveOrDie
I dont care I use intel AMD sucks.
Posted on Reply
#53
Raiden85
LiveOrDieI dont care I use intel AMD sucks.
If it wasn't for AMD kicking Intel up the ass we would stiil be stuck with quad cores and the overpriced HEDT platform. About 10yrs of quads was enough.
Posted on Reply
#54
R-T-B
Gentlemen, keep your fanboy at the thread entrance...
Posted on Reply
#55
Unregistered
R-T-BGentlemen, keep your fanboy at the thread entrance...
Yeah there's enough AMD ones in TPU now anyway.
Posted on Edit | Reply
#56
R-T-B
TiggerYeah there's enough AMD ones in TPU now anyway.
Brand don't matter, just stow it if that's all you got.
Posted on Reply
#57
Unregistered
R-T-BBrand don't matter, just stow it if that's all you got.
I never unstowed it did i
Posted on Edit | Reply
#58
R-T-B
TiggerI never unstowed it did i
Wasn't talking about you... but uh... I guess?
Posted on Reply
#59
Unregistered
R-T-BWasn't talking about you... but uh... I guess?
You was you replied to me, Brand don't matter, just stow it if that's all you got. But no worries eh
Posted on Edit | Reply
#60
R-T-B
TiggerYou was you replied to me, Brand don't matter, just stow it if that's all you got. But no worries eh
I was speaking generally in response to the AMD part. But no worries yeah.
Posted on Reply
#61
Patriot
fiftofarPersonally never had a good time with AMD drivers in general, when Windows updates are thrown in the mix it is even worse (mainly with alpha/beta/insider etc)
Especially the AMD RAID drivers and software.
Pre-nvme their raid drivers were fine...their testing of nvme+hdd raid was lackluster and a pita to use.
Now in combination with windows update touching drivers they shouldn't...
Needless to say I moved that friend to storage spaces... I figure they are more likely to test compatibility with things they make.

Windows has changed driver frameworks once or twice with windows 10 and I have seen many driver issues due to that from many vendors. There is no longer windows 10 compatibility, there is a recipe of Win10build +driver.
AMD gpu drivers have been more or less unstable trying to use their screen recording capabilities, but then again shadowplay has gone through bouts of instability.

@bug, AMD having to make a new driver seems more like another framework change with the changes Intel made to the scheduler, Keep in mind Intel and MS made the scheduler for win11 together.
Posted on Reply
#62
TheoneandonlyMrK
PatriotPre-nvme their raid drivers were fine.

Now in combination with windows update touching driers they shouldn't...
Needless to say I moved that friend to storage spaces... I figure they are more likely to test compatibility with things they make.
I've used raid0 on Amd for years across builds and HDD and nvme without a single issue.

Hopefully you will continue to be wrong in the future but I'll let you know if I ever agree with that.

Now on Topic.

Good a new OS tends to have teething problems good to see prompt fixes.
Posted on Reply
#63
Patriot
TheoneandonlyMrKI've used raid0 on Amd for years across builds and HDD and nvme without a single issue.

Hopefully you will continue to be wrong in the future but I'll let you know if I ever agree with that.

Now on Topic.

Good a new OS tends to have teething problems good to see prompt fixes.
Wow toxic trash.

Windows changes driver frameworks, it causes driver issues. Thems the facts.

If you don't keep your driver in lockstep recipe it causes issues.
Windows 11 changed stuff again, and Intel helped write the scheduler hence new driver required.

AMD had issues with young Windows 10 drivers and nvme+raid. I wrote multiple bug reports and got them fixed...
Having to have 3 different drivers on install to enable hdd raid from nvme boot... is still a jank install method.

Whose fault is it? Well windows 10 has been a bit of a moving target for driver writing...
Drivers written on earlier frameworks are supposed to work on later but MS has broken support between "compatible" versions multiple times.

Windows breaks stuff with updates, and vendors are forced to find fixes and release drivers that work around MS changes.
Posted on Reply
#64
TheoneandonlyMrK
PatriotWow toxic trash.

Windows changes driver frameworks, it causes driver issues. Thems the facts.

If you don't keep your driver in lockstep recipe it causes issues.
Windows 11 changed stuff again, and Intel helped write the scheduler hence new driver required.

AMD had issues with young Windows 10 drivers and nvme+raid. I wrote multiple bug reports and got them fixed...
Having to have 3 different drivers on install to enable hdd raid from nvme boot... is still a jank install method.

Whose fault is it? Well windows 10 has been a bit of a moving target for driver writing...
Drivers written on earlier frameworks are supposed to work on later but MS has broken support between "compatible" versions multiple times.

Windows breaks stuff with updates, and vendors are forced to find fixes and release drivers that work around MS changes.
Yeah and yet still no problems here hmmn.
Posted on Reply
#65
bug
PatriotPre-nvme their raid drivers were fine...their testing of nvme+hdd raid was lackluster and a pita to use.
Now in combination with windows update touching drivers they shouldn't...
Needless to say I moved that friend to storage spaces... I figure they are more likely to test compatibility with things they make.

Windows has changed driver frameworks once or twice with windows 10 and I have seen many driver issues due to that from many vendors. There is no longer windows 10 compatibility, there is a recipe of Win10build +driver.
AMD gpu drivers have been more or less unstable trying to use their screen recording capabilities, but then again shadowplay has gone through bouts of instability.

@bug, AMD having to make a new driver seems more like another framework change with the changes Intel made to the scheduler, Keep in mind Intel and MS made the scheduler for win11 together.
I doubt the new scheduler was MS and Intel only, AMD was most likely involved too. And let's not forget Qualcomm. The bug was linked here at some point, it's 6 months old. So it wasn't a last minute change, bat rather something both MS and AMD were ok with - they might have deemed it not critical or agreed on a timeline beforehand. Either that, or they spat about behind the curtains.
Posted on Reply
#66
R-T-B
TheoneandonlyMrKYeah and yet still no problems here hmmn.
Luck you.
Posted on Reply
#67
Patriot
TheoneandonlyMrKYeah and yet still no problems here hmmn.
I am glad you are enjoying the results of my bug reports, subsequent patches, and updated install directions, you are welcome.
bugI doubt the new scheduler was MS and Intel only, AMD was most likely involved too. And let's not forget Qualcomm. The bug was linked here at some point, it's 6 months old. So it wasn't a last minute change, bat rather something both MS and AMD were ok with - they might have deemed it not critical or agreed on a timeline beforehand. Either that, or they spat about behind the curtains.
Main point, Dev build had the fix without new AMD driver. I am sure the agreed on schedule had a Fix before RTM not after... and MS may have released earlier than originally scheduled to correspond with ADL launch. I have seen plenty of companies having a good face together and fighting behind curtains.
Posted on Reply
#68
bug
PatriotMain point, Dev build had the fix without new AMD driver. I am sure the agreed on schedule had a Fix before RTM not after... and MS may have released earlier than originally scheduled to correspond with ADL launch. I have seen plenty of companies having a good face together and fighting behind curtains.
Which fix is that? Because the cache latency seems to have been fixed by MS, but that preferred core was AMD's to fix.
And while I'm glad you're sure about what happened between MS and AMD, I'm not and as such I choose not to point fingers.
Posted on Reply
#69
TheoneandonlyMrK
PatriotI am glad you are enjoying the results of my bug reports, subsequent patches, and updated install directions, you are welcome.


Main point, Dev build had the fix without new AMD driver. I am sure the agreed on schedule had a Fix before RTM not after... and MS may have released earlier than originally scheduled to correspond with ADL launch. I have seen plenty of companies having a good face together and fighting behind curtains.
Thank you lord but I did say I've been using it since FX days ,was that you fixing the issues I wasn't having back then too, praise to you bro.
Posted on Reply
#70
Patriot
TheoneandonlyMrKThank you lord but I did say I've been using it since FX days ,was that you fixing the issues I wasn't having back then too, praise to you bro.
Reading is hard.
Posted on Reply
#71
ValenOne
It looks like AMD Ryzen Chipset Driver 3.10.08.506 doesn't support AMD Ryzen Mobile 2000U Series with Windows 11.
Posted on Reply
#72
Patriot
bugWhich fix is that? Because the cache latency seems to have been fixed by MS, but that preferred core was AMD's to fix.
And while I'm glad you're sure about what happened between MS and AMD, I'm not and as such I choose not to point fingers.
Why was it AMDs to fix? The frameworks are supposed to be backwards compatible, yet oddly often require new drivers from...everyone. (shrugs)
And I am sure they agreed on a fix, since they listed timelines on their (amd) website... Every OS before 10 had 1 framework listed... 10 has.. 10, and 11 yet another.
They are all supposed to be backwards compatible yet the windows 10 driver doesn't have preferred core working on 11, why is that?

My main point has been its the constant changing frameworks that are causing the driver headaches all around. Drivers built on older builds may work on newer or they might not... it's been pretty hit and miss on windows 10 rolling release.

I am not saying I know what is going on behind the scenes... Just saying those framework changes are not as compatible as they claim, and why would they be, they can't test everything, they are just being reckless at the speed of changes.
Posted on Reply
#73
Mussels
Freshwater Moderator
LiveOrDieI dont care I use intel AMD sucks.
Shhh, you know better.
rvalenciaIt looks like AMD Ryzen Chipset Driver 3.10.08.506 doesn't support AMD Ryzen Mobile 2000U Series with Windows 11.
That's because the 2000U series are Zen1 chips, unsupported in 11.
Posted on Reply
Add your own comment
Nov 23rd, 2024 23:12 EST change timezone

Latest GPU Drivers

New Forum Posts

Popular Reviews

Controversial News Posts