x86-64 box

Christopher Browne cbbrowne-Re5JQEeQqe8AvxtiuMwx3w at public.gmane.org
Sun Jun 12 21:00:03 UTC 2005


On 6/9/05, Ken O. Burtch <kburtch-Zd07PnzKK1IAvxtiuMwx3w at public.gmane.org> wrote:
> On Thu, 2005-06-09 at 22:40, William Park wrote:
> > On Wed, Jun 08, 2005 at 02:40:23PM +0800, JM wrote:
> > > hi,
> > > 
> > > 	we will be buying a 64-bit box, i was wondering if i need to
> > > 	recompile my kernel for this?
> > 
> > Answer is "No".  Speaking from i386 vs. i686 experience, I barely
> > noticed the difference.  Screen updates were faster, but not by that
> > much.
> 
> I don't know if this has been mentioned, but you'd probably get a bigger
> performance boost by using a SMP kernel rather than recompiling a
> non-SMP kernel for AMD.  SMP (Symmetric Multiprocessing) will take
> advantage of hyperthreaded processors and you should always install an
> SMP kernel on a modern computer.

Exqueeze me?

1.  It is highly debatable whether or not hyperthreading provides
*any* improvement.  The "best" cases are estimated at a 20%
improvement, and that assumes that the multiprocessing doesn't lead to
cache poisoning that would worsen things.

For the most part, the Xeon architectures wind up with memory
throttling that outweighs such improvements...

2.  AMD has never implemented hyperthreading on their processors.  As
a result, this isn't a relevant optimization for AMD-based systems.

3.  SMP will only be relevant if you actually have more than one AMD
CPU on your system.
-- 
http://www3.sympatico.ca/cbbrowne/linux.html
"The true  measure of a  man is how he treats  someone who can  do him
absolutely no good." -- Samuel Johnson, lexicographer (1709-1784)
--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml





More information about the Legacy mailing list