Bulldozer drivers or Windows 7 hotfix?

[keyser]

AMD also shared with us that Windows 7 isn't really all that optimized for Bulldozer. Given AMD's unique multi-core module architecture, the OS scheduler needs to know when to place threads on a single module (with shared caches) vs. on separate modules with dedicated caches. Windows 7's scheduler isn't aware of Bulldozer's architecture and as a result sort of places threads wherever it sees fit, regardless of optimal placement. Windows 8 is expected to correct this, however given the short lead time on Bulldozer reviews we weren't able to do much experimenting with Windows 8 performance on the platform. There's also the fact that Windows 8 isn't expected out until the end of next year, at which point we'll likely see an upgraded successor to Bulldozer.

http://www.anandtech.com/show/4955/the-bulldozer-review-amd-fx8150-tested/11

Maybe Bulldozer would benefit from some drivers for Windows 7 or we could get some future tech changes now from Microsoft in the form of hotfixes? Discuss

 

[CAT-THE-FIFTH]

Toms Hardware tested the FX8150 under the developers preview of Windows 8:

http://www.tomshardware.co.uk/fx-8150-zambezi-bulldozer-990fx,review-32295-23.html

In multi-threaded applications there seems to be no difference. However,in WoW performance increased by around 10% at 1680X1050 and 1920X1080 and on top of this power consumption dropped by 10% at idle. WoW only uses two threads AFAIK.

OTH,there is no guarantee ATM there will a patch for Windows 7 so performance will be where it is in current reviews. Of course the will better application optimisations for the FX8150 as time progresses but these also probably affect Intel CPUs as much as AMD ones I suspect.

 

[Sin_Chase]

I'm calling it - Single figure % average increase with a scheduler fix.

When the Chip is fully loaded and utilising ALL resources I do not see how more intelligent thread assignment can help anything. It's all in use anyway.

 

[DragonQ]

I'm calling it - Single figure % average increase with a scheduler fix.

Someone on another forum did a test with all four modules only running one core and, in some cases at least, the performance increase was well into double figures. Whether an improved scheduler can do quite as well is anyone's guess.

When the Chip is fully loaded and utilising ALL resources I do not see how more intelligent thread assignment can help anything. It's all in use anyway.

True.