Microsoft Releases Hotfix to Improve Bulldozer Performance UPDATE: Pulledby Kristian Vättö on December 16, 2011 9:29 AM EST
The 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.
Post Your CommentPlease log in or sign up to comment.
View All Comments
jeremyshaw - Friday, December 16, 2011 - linkIMO, that's what they have done in the past, slowly, deliberately, and surefootedly, marched intels pace, even if intel itself failed to keep up. Though they did try something new: APU. While we may not think much of it, it is apparently running the roost in some markets outside of the USA.
The manuf aspect of the technology race? I honestly don't expect anyone to catch up to intel in the consumer market. iBM may give them a helluva chase (and whooping) in the enterprise/server market, though (technically, it's what's happening right now).
0ldman79 - Saturday, December 17, 2011 - linkThere were several players a few years ago. Now there is only AMD and Intel. Anyone else (Via?) is left to ultra low power usage with matching performance.
Most of the x86 compatible single board computers are AMD based.
Otherwise, I agree, AMD needs to up the ante a bit. They had the performance crown for a while, now they're pricing is much more agreeable and the CPU's are fast enough, but we now have more of a performance difference percentage than we did back in the P2/K6 days.
chizow - Friday, December 16, 2011 - linkThis hotfix doesn't do anything to address Bulldozer's underlying problem, poor design leading to poor single/lightly threaded performance. All it does is arrange thread load on the fewest number of modules to give a slight boost in performance from turbo modes for lightly threaded apps. Given how far behind Bulldozer is compared to Intel or even previous Phenom II designs, don't expect any miracles from a few hundred MHz.
JonnyDough - Saturday, December 17, 2011 - linkSpoken like a true Intel fangirl.
0ldman79 - Saturday, December 17, 2011 - linkI don't see a bit of preference in that comment. Sounds like a bit of fact.
Bulldozer is very reminiscent of Williamette. First gen of a new core that needs GHz+ to work well compared to the previous generation.
I was kind of waiting of Bulldozer to start my next system, now I'm really leaning towards a 6 core Phenom II if it will work in my board, or a faster quad core if it won't.
Completely ignoring Intel, Bulldozer performance just isn't that impressive compared to the previous generation at this time.
I'd rather have the six core CPU.
chizow - Saturday, December 17, 2011 - linkRight, coming from the gal dancing around this thread defending this dud of a CPU at every chance.
Its a dud, even AMD has admitted to it and are looking to correct their mistake ASAP with Piledriver.
Hector2 - Friday, December 16, 2011 - linkI'm confused. I'll take a leap of faith that AMD validates their products before launch and that they would include Microsoft OS'es in their validation. Did AMD wait until Anand & Co ran their benchmarks and uncovered a performance problem ?
Seems like Elementary Validation 101 to me. If they already knew they needed a Microsoft fix, they should have either 1) delayed launch until it was fixed or 2) at least mention that we would see performance degradation until it was fixed by MS.
This shouldn't have been a surprise to either AMD or their customers.
Kristian Vättö - Friday, December 16, 2011 - linkThey did the latter one. Like I mentioned in the article, AMD ran tests on W7 and W8 before the NDA of Bulldozer was lifted, so they knew what was going on.
BSMonitor - Friday, December 16, 2011 - linkIt is not a fix, it's an optimization. And Microsoft is under no obligation to optimize their code for every AMD whim out there.
This all sounds a lot like 3D Now!.
Intel builds and implements its own compilers when implementing new registers/architecture changes, etc. And hands that over to the software companies.
AMD makes hardware changes and expects everyone to do the work for them.
Two different approaches. Clearly one is superior.
aGreenAgent - Friday, December 16, 2011 - linkIntel does the same thing. Look at the number of Windows KBs that fix (in software) a hardware problem that Intel did. This is fairly standard practice for all hardware manufacturers with respect to Windows fixes.