Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 9 Feb 1999 10:13:10 -0700 (MST)
From:      Kevin Van Maren <vanmaren@fast.cs.utah.edu>
To:        Thierry.Herbelot@alcatel.fr, smp@FreeBSD.ORG
Subject:   Re:  Mixing P2 Types
Message-ID:  <199902091713.KAA17064@fast.cs.utah.edu>

next in thread | raw e-mail | index | archive | help
Well, it *should* work, but they will both have to run at the
speed of the slowest processor.  The only problem with that is
Intel has added multiplier-locking to many of the Pentium II
processors, preventing them from running if the multiplier is
wrong.  If they just locked the multiplier, you would have
processors at different speeks (probably okay), but they
won't POST if the multiplier is changed.  Allegedly the
multiplier isn't locked for `older' P-IIs with the locking if
they are running at 66MHz bus.

There are a few high-end machines that supported unequal Pentium
processors ($80k+ range), and some of the Quad Pro boards could
set the multiplier for the pairs of CPUs independantly, but I
don't know of any Pentium II board that will do that.

As for the SMP spec, that is true.  However, Intel has said
that they will not support mixing speeds/caches on the Xeon
line, although (I believe) Compaq has said that they will.

It it wasn't for the multiplier locking on the newer P-IIs,
I'd tell you to get a 350 (or faster) CPU and the P2B-DS, and
run the 350 at 266 until you upgraded to a second 350.  The
faster bus makes a bigger difference under SMP due to increased
usage of the shared bus, and the 350 isn't much more than the 333.
Maybe getting two 350s isn't too far out of reach?

Kevin

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-smp" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199902091713.KAA17064>