Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!uakari.primate.wisc.edu!dogie.macc.wisc.edu!uwvax!ub.d.umn.edu!pbye
From: pbye@ub.d.umn.edu (peter bye)
Newsgroups: comp.benchmarks
Subject: AT&T StarServer E Perfomance.
Message-ID: <570@ub.d.umn.edu>
Date: 14 Dec 90 17:22:16 GMT
Sender: pbye@ub.d.umn.edu
Organization: University of Minnesota, Duluth. Information Services.
Lines: 33

I don't suppose anybody has any of the benchmarks for AT&T's StarServer E
Symetric Mutiprocessor box yet, in particular "bc".

I have a few that I will share -- someone gave them to me with out
restriction, so I see no reason why you can't see them also.

Machine                     KDrystones*

AT&T 6386/33                   17
SPARCstation 300               28
IBM 3090/200                   31
StarServer E (one processor)   47


Machine                     SPECmark:    Geometric Mean of Integer results 

StarServer E (one processor)  11.6                       17.1
RS/6000 520                   22.3                       15.8
SPARCstation 330              11.3                       12.3 


Machine                         TP1 Benchmark

Sequent Symetry (8-processors)      82
IBM 3090/400                        62  Transactions per second.
StarServer E (one processor)        53


If anyone can tell me were to get the code for bc, I see that it gets run on
a StarServer, and publish the results.

Peter Bye
pbye@ub.d.umn.edu