On Fri, 5 Jun 1998, Bradley, Greg wrote:
>We have tested our process with a single celeron and get the required
>performance, but have two questions:
> 1/ Will 2 celerons work together, we can't think of a reason why
>not but thought we should ask
Just got out of the Intel Processor Dealer training where this question
was asked. The Intel rep said NO. The Celerons will not work in any dual
processor configurations. They will fit in the Slot 1 of a dual
processor board like the N440BX (if you have the right retention
mechanisms which this board lacks), but will not work. (No lines for
multiprocessor signalling apparently) The training manual on the Celeron
says the same thing as well. It only works on uniprocessor designs.
> 2/ Will we get the same performance gain from a celeron pair as
>we would from a PII pair or would they be competing to much for access
>to main memory without the cache's.
Because the Celeron has no L2 cache it will be significantly slower. The
official figure is that a true P-II will be 25% faster than an equally
clocked Celeron. (For only about $100 saving on system price)
We will be running 100MHz
>motherboards so we think the smaller difference between cpu and main
>memory speeds will offset this to some extent.
Celerons only work at 66MHz system bus speeds. There are no plans
currently for 100Mhz.
>Has anyone tried this and can they give us their feelings on the matter
>(and any stats they may have gathered :-) ).
If you get the impression that the Celeron is a dain-bramaged idea...
You are right.
Stephen Zedalis
System Administrator - Exis Net, Inc.
--
PLEASE read the Red Hat FAQ, Tips, Errata and the MAILING LIST ARCHIVES!
http://www.redhat.com/RedHat-FAQ /RedHat-Errata /RedHat-Tips /mailing-lists
To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject.
