Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!asuvax!ncar!elroy.jpl.nasa.gov!swrinde!mips!spool.mu.edu!cs.umn.edu!umnstat.stat.umn.edu!rjg
From: rjg@umnstat.stat.umn.edu (Robert J. Granvin)
Newsgroups: comp.sys.hp
Subject: "Demo" 720's vs. real ones.
Summary: Performance of pre-release 8.0 vs. released 8.0?
Message-ID: <1991Jun13.160938.7177@cs.umn.edu>
Date: 13 Jun 91 16:09:38 GMT
Sender: news@cs.umn.edu (News administrator)
Reply-To: rjg@umnstat.stat.umn.edu (Robert J. Granvin)
Organization: University of Minnesota, School of Statistics
Lines: 27
Nntp-Posting-Host: itasca.stat.umn.edu

Has anyone had the opportunity to demo an HP 720 and then use a "Real"
one?  

My area of interest is in what differences may have been noticed.  It
appears that a number of demo machines are running pre-release (beta)
versions of HP/UX 8.0 and do not include the good optimizer compilers.

With only about 48 hours to work with a unit, we did some benchmark
testing.  While we found the machine impressive, it did seem that the
graphics performance _should_ have been better.  (Certainly we can 
overlook the obvious bugs in a pre-release OS :-).  The 2-D acceleration
seemed to give us a 3x boost in performance which can be explained, at
least in part, by a much faster CPU.  A question of curiosity is whether
the accelerator was actually being used to its full extend by the pre-
release system (or at all).

Until we can get to a "real" 720 with released software, it would be
interesting to hear if anyone has had the opportunity to compare...
Does the release OS/compilers make any real difference on performance
(in any categories, graphics specifically...)  Any HP techs online
that can assist?

email, please.

-- 
Robert J. Granvin                                       School of Statistics
rjg@umnstat.stat.umn.edu                             University of Minnesota