Bulldozer Windows 7 Hotfix

[Cazalan]

Let the testing commence!

http://support.microsoft.com/kb/2592546/

 

[BaronMatrix]

Let the testing commence!

http://support.microsoft.com/kb/2592546/

Let's see how long it take for the naysayers to make excuses.

 

[4745454b]

Let's see how long it take for the naysayers to make excuses.

Better question is what does this actually do? By how much does everyones benchies go up? If this changes the games, then great for AMD and us. If it adds another 1-5%, who cares? If BD is ~88-90% slower then PhII, then another 3% only gets them close to where they already were.

 

[loneninja]

Let's see how long it take for the naysayers to make excuses.

Better question is what does this actually do? By how much does everyones benchies go up? If this changes the games, then great for AMD and us. If it adds another 1-5%, who cares? If BD is ~88-90% slower then PhII, then another 3% only gets them close to where they already were.

Toms did a preview of Win8 performance and the Win8 scheduler for Bulldozer, the Win7 patch here should be the same thing. Performance gains were from allowing more threads to run on the same module so that unused modules could be disabled allowing the active modules to turbo higher. Performance gains were there, but they weren't great. This patch should have zero impact on fully threaded work loads or intense multitasking, or for the overclocker with turbo disabled. But now that guy running around this forum claiming a 20% performance gain can finally be proven wrong. :lol:

 

[4745454b]

I was thinking the same thing. Not sure its as easy as giving win7 the same scheduler as win8 however. It would be nice to see some benchies however so we have a clearer picture as to what it really does.

Edit: Anand link. http://www.anandtech.com/show/5251/microsoft-releases-hotfix-to-improve-bulldozer-performance

We have not yet tested Bulldozer with the hotfix, but don't expect miracles. Better scheduling for the Bulldozer CPUs will improve performance a bit, but not enough to close the gap in many scenarios. Windows 8 already has the new thread scheduler and according to AMD's own and third party tests, the performance increase is up to around 10%.

So if you get another 10%(ish?) and BD is ~88-90% slower then PhII, then this means you are looking at performance around the PhII? Give or take a bit? And assuming you have the thermal room for the cores to ramp up? Good job AMD/MS for getting a hotfix to increase performance. But I think we need a bit more. Hopefully Piledriver can do it.

 

[Cazalan]

They're gonna need to fix it pretty soon. The new Llano or trinity CPUs which use the Bulldozer+ cores are supposed to be out in January. They need every bit of performance they can get since it's stuck with a 2 module (FX4100 style) CPU.

 

[Mousemonkey]

Let's see how long it take for the naysayers to make excuses.

Well it's already been removed from the MS website, so it's not the naysayers who are having to make excuses.

http://www.pcper.com/news/Processors/Microsoft-Releases-Update-Improve-Bulldozer-I-thought

 

[phatbuddha79]

Let's see how long it take for the naysayers to make excuses.

Even if BD improves 10-15%, will it also improve its power draw? The 2500k can be had for less, currently stomps it in most apps, and far more efficient.

 

[noob2222]

However, less than 24 hours after introduction, Microsoft pulled the patch. We contacted AMD and Microsoft and received an answer from AMD first. For starters, AMD was surprised that Microsoft decided to post the patch in its current form, because:


"there are actually two updates needed for AMD Bulldozer CPU architecture. Microsoft posted just the first patch and we do not believe users would benefit in any way from it. The patch was originally scheduled for the first quarter 2012 and then the users will see tangible performance benefits when using Windows 7 and Windows Server 2008 R2 operating systems."


- BSN

 

[ElMoIsEviL]

http://ht4u.net/news/24857_patch_soll_bulldozer_unter_windows_7_beschleunigen_-_ergebnisse_enttaeuschen/

It is a mixed bag. So we the "naysayers" get to tell you the Propagandist where you can go once more.

Tired of all this misinformation. And no... a re-spin alone will not make up the performance deficit. There needs to be some fundamental changes made to the core architecture. Mostly the branch prediction, caching mechanism and obviously more FPU power (at least until OpenCL/DirectCompute become mainstream). Then it would be competitive... not on a per clk basis but as a product overall. Swallow thy pride and stop this ridiculous debate Baron.

 

[4745454b]

It is pulled. AMD said its only half the problem, they are working on the other half. MS seems to have jumped the gun a hair...