XP Performance

Jonathan Hutchins hutchins at opus1.com
Thu Nov 29 18:25:38 CST 2001


We were discussing this a while back.  Still not many ideas about
benchmarking XP against Mandrake 8.1.

InfoWorld's notorious test showing W2K is faster than XP:

http://www.infoworld.com/articles/tc/xml/01/10/29/011029tcwinxp.xml

They used a 1.5G P4 system and a 700MHz system.  The slower system showed
the differences better.  Given that the newest machines in my office are
1.0GHz/128M RAM, and more than half are 400MHz or less, the older hardware
is more realistic.  We've just finished removing and upgrading all systems
that were three years old or older.

Tech-report.com did some extensive testing with 1.5G and 800MHz AMD
Athalons, 512 and 256M RAM:

http://www.tech-report.com/reviews/2001q4/os/index.x?pg=1

 With what I would consider very current high-end and mid-range systems,
they were able to show XP to be faster than W2K and ME.  I think what we're
seeing there is fast hardware overcoming the code bloat.

What I'd really like to see would be a mid-range system like an Athalon 800
against a realistic system from my office - either a new-this-year P3 600 or
a two-year-old PII 400.  Sure, give 'em all a RAM upgrade to 256M, it's
cheap enough these days that even we would spring for it if we were
upgrading (which we're NOT).  And I'd like to see XP compared to NT4, which
we're actually running, rather than W2K which is a lot closer in code to XP.
I doubt all that many companies are going to upgrade to W2K, then
immediately upgrade to XP before the license fees are amortized (3 years).

In the TR Comments section there are a number of supportive comments that
say XP has benchmarked slower in the real world as well.

Jonathan




More information about the Kclug mailing list