Hello, been a while, dual CPU mobos

Lennart Sorensen lsorense-1wCw9BSqJbv44Nm34jS7GywD8/FfD2ys at public.gmane.org
Tue Sep 7 03:27:21 UTC 2004


On Mon, Sep 06, 2004 at 10:13:36PM -0400, Christopher Browne wrote:
[snip]
> We've got quad-Xeon boxes at work that actually start to _slow down_ as
> you scale up assortedly because:
> 
>   a) Hyperthreading pretty much just sucks;

That is not necesarily true.  Intel's implementation may be the problem
(like many other parts of the P4 design which just seems full of strange
bottlenecks).

>   b) Context switches get _way_ more expensive when you break the
>      2GB barrier, which means that the last 6GB of memory on these
>      boxes actually _slow_ performance, despite the stunning cost
>      of the DIMMs :-(
 
Well that is certainly a big one.  PAE's memory mapping or banking
certainly doesn't help speed up things, although it should still beat
swap space.

> The big, big win with AMD-64 is _not_ primarily in the ability to have
> more CPUs, although it looks like it's no slouch there; the win is that
> you can address 32GB of RAM without the performance-destroying hackery
> of the Intel Xeons.

The athlon 64s also seem to get a lot more done per clock than the Xeons
do.  And they actually seem to be designed to work as a 64bit chip.

Lennart Sorensen
--
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