[ILUG] NT Vs Linux Benchmark Part 2, the return of NT?????
drjolt+ilug at redbrick.dcu.ie
Mon Nov 15 13:25:10 GMT 1999
Quoting <383005DA.95E2100D at online.ie>
by "John P. Looney" <jplooney-mutt at online.ie>:
> Robert Sweetnam wrote:
> > The second benchmark results are available at the following
> > http://www.mindcraft.com/whitepapers/openbench1.html A bit fairer
> > this time but the results are similar, Still I think I'd prefer
> > stability over negligible performance gains.
> I couldn't but note that "June 30,1999" was up at the top of the
> document. It was used by a lot of linux heads, to start looking for
> performance bottlenecks. Hopefully, we'll see all this work in a few
> months - when 2.4 comes out, and when the application use the new
> hooks it introduces...
Although I only skimmed it, the impression I got from reading it was
"Neither OS demonstrated much in the way of MP scalability" there's a
quote in there that a 4-way NT system provided a 107% performance gain
over a 1-way NT system - I'm pretty sure a 4-way Sun does a trifle
better than twice the performance of a 1-way system!
"You can ISO9001 certify the process of shooting yourself in the foot,
so long as the process is documented and reliably produces the proper
result." - Adam Shostack, on bugtraq
David Murphy - For PGP public key, send mail with Subject: send-pgp-key
More information about the ILUG