Microsoft Releases Hotfix to Improve Bulldozer Performance UPDATE: Pulled
by Kristian Vättö on December 16, 2011 9:29 AM ESTThe launch of Bulldozer in October wasn't exactly a success for AMD. In our review, Anand ended up recommending the Intel i5-2500K over AMD FX-8150. One of the reasons behind the poor performance of Bulldozer is its unique design: each Bulldozer module consists of two integer and one floating point core. Todays operating systems don't know how to optimally schedule threads for this design and as a result, the full potential of Bulldozer has not been achieved. Microsoft has released a hotfix for Windows 7 and Server 2008 R2 that should increase the performance of Bulldozer.
Let's look at the problem to see what happened and how the hotfix helps address it. Before the update, Windows didn't know how to ideally schedule threads on Bulldozer. Essentially, it didn't know when it was good to place threads on single module versus multiple modules.
The picture above explains this pretty well. Before the update, Windows more or less randomly placed the threads which meant many modules were unnecessarily active at the same time. This capped the maximum Turbo speeds because those can only be achieved when some of the modules are inactive (power gated).
VR-Zone is claiming that Windows sees one Bulldozer module as a single multi-threaded core, similar to an Intel Hyper-Threading core. Basically, your 8-core FX-8150 is seen as a quad-core, 8-thread CPU—just like Intel's i7-2600K for instance. This goes against AMD's design and marketing because Bulldozer is closer to an 8-core CPU.
We have not yet tested Bulldozer with the hotfix, but don't expect miracles as Microsoft is suggesting a 2-7% increase. 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%, but Bulldozer needs a lot more than 10% to surpass Sandy Bridge.
Update: VR-Zone reports (and we can confirm) that the download link for the hotfix is no longer functional. There were apparently unexpected performance drops in some cases after applying the hotfix and Microsoft is investigating the issues. Modifying the scheduler in Windows is not something to be done lightly, as it changes a core element of the OS, so more testing and validation for such updates is always a good idea.
Update 2: Apparently there is a second part to the hotfix that was not pushed live, and this hotfix was pushed live prematurely.
Source: Microsoft
Facebook
Twitter
RSS
75 Comments
View All Comments
BSMonitor - Friday, December 16, 2011 - link
Again, it is not a fix. It is optimizing software for a architecture enhancement. Similar to Hyperthreading, Quicksync, MMX, SSE, etc..If AMD wants software designers to make full use of <new> transistors, AMD needs to provide the tools for doing this. Not wait around for software vendors to incorporate their new tech.
AMD has compiler code available to it as well as Intel... Intel simply does this better than AMD.
Again, NOT a fix. An Optimization.
phatboye - Friday, December 16, 2011 - link
This is an optimization to the Windows scheduler, not some library optimization. This is nothing like MMX, SSE or quicksync. Neither AMD nor Intel has the ablility to release updated versions of the windows scheduler as that is proprietary MS code. They have no access to that and only MS is able to produce optimizations to that code.BSMonitor - Friday, December 16, 2011 - link
Yawn. You must be a Republican. Nit pick the evidence. And miss the point completely.Notice how you pulled out the three similar scenarios and left the one scenario that is EXACTLY like this out of your argument.
Good try, no one is saying what you are implying I said. Grow Up.
phatboye - Friday, December 16, 2011 - link
OK, I get it. Of the 4 things you mention 3 of them have nothing to do with the topic at hand and I call you out on it. Then your response was that I am nick picking. Finally you resort to calling me a Republican (lol). You then conclude by telling me to grow up while showing your level of maturity (which is not that of an adult).OK yeah I left out hyperthreading. Even still Intel does not have access to optimize the Windows scheduler to improve hyperthreading performance so my argument still stands. You cannot blame AMD for this as this is something totally out of their hands. Even if AMD wanted to release an optimized scheduler they can't. I wound't go around and say MS is at fault either, maybe they wanted to wait till final silicon was released in order to properly test this optimization out. Maybe they needed more time for validation. Either way my argument still stands.
JonnyDough - Saturday, December 17, 2011 - link
I agree with your "immature" statement. /threadBSMonitor - Monday, December 19, 2011 - link
My immature reaction is entirely eaggerated to meet the level of ignorant ranting of the person commenting.Continues to fail at comprehending the initial point and repeats his own beliefs, instead of facts.... YAWN
phatboye - Monday, December 19, 2011 - link
LOL ignorant ranting? You initial point was that AMD was not providing the fix themself and instead is relying on others (MS in this case) to provide a fix for their changed hardware.I then explained how this is something that is totally out of their control. MS is the only one able to make changes to the MS scheduler.
You have yet to counter my argument with anything (substantial or otherwise) at all. Instead you question my maturity and resort to stating that I "Continues to fail at comprehending the initial point".
The really funny part is when you state that I am ignorant on this subject. That is laugh worthy. While AMD doesn't provide it's own proprietary compiler and no where near as much code is released from AMD as it is from Intel (AMD doesn't have nearly as many resources as Intel) to state that AMD does nothing is completely false. It is just as ignorant as your claim that AMD should do the work of optimizing the scheduler themselves especailly when Intel does the same exact thing when they release a new CPU architecture requiring core changes to the MS windows closed source scheduler.
Also consider this my last update to your post unless you are going to post something relevant to defend your original statement. Otherwise you are just trolling with your claims of me being ignorant without any counter-argument to back up your statements.
JonnyDough - Saturday, December 17, 2011 - link
Stop calling it that. Windows needs to change, because the architecture is FORWARD THINKING and cannot be changed. Its a FIX, something that has to be implemented to make the hardware work properly. You really are an Intel fanboy, I bet you own stock...You can argue semantics all day long if you'd like, but its kind of silly. Who cares what its called?
BSMonitor - Monday, December 19, 2011 - link
Intelligent human beings.Fix = implies Microsoft has done something wrong in not preemptively changing long-standing base code for a 3rd parties' architectural whim
Optimization = a planned redesign of existing code to accommodate new hardware
My point, AMD waits until after the release, and then blames Microsoft.
Intel works with Microsoft during development to ensure optimization at launch.
Again, ignorant fools prove their amazingness on forums!
JonnyDough - Saturday, December 17, 2011 - link
You're right. They should just design their operating system around Intel and give them a monopoly. /sarcasm