DragonFly BSD
DragonFly users List (threaded) for 2005-02
[Date Prev][Date Next]  [Thread Prev][Thread Next]  [Date Index][Thread Index]

Re: Dragonfly and Hyperthreading....

From: Tom Hummel <tom@xxxxxxxxxxxxx>
Date: Mon, 21 Feb 2005 21:40:52 +0100

    is that performance is extremely assymetric.  You could waste hundreds
    of man hours writing an HT aware scheduler and still not have it work
    well for all situations.

Microsoft implemented in a way that made many people happy, and HT had it's impact on the minds of many. There should be some reward, but i don't know what impact such implementation could have on the rest of the code. Could you give us some details?

Even more to the point, HT technology was most useful in past years
when instruction pipelines were not able to make full use of the cpu's
resources. That has changed as well. Today's cpu's instruction pipelines are able to utility a far higher percentage of the cpu's
available resources, making HT a lot less useful. HT has only been
AFAIR Intel intentionally slowed down pipeline operations to the execution units, so they could push up the clock frequency, to show customers some really high numbers, and later (2004) they discovered netburst isn't going to go to 10Ghz, as they predicted in 2003.
I think this is an Intel specific mistake, but nonetheless makes SMT useless on cpus with well designed pipelines.

viable up till now because Intel's cpu architecture sucks rocks compared
with AMD's, but Intel can no longer compete by boosting clock rate so
now they are stuck... they have to slow their cpus down and make them
more efficient, and that will kill HT's effectiveness even if you ignore the heat issue.
Maybe HT will be the only way to gain more effectiveness with lower clock frequency on netburst (P4).
My point of interested is what's AMD making better?

. ..tom

[Date Prev][Date Next]  [Thread Prev][Thread Next]  [Date Index][Thread Index]