Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site topaz.ARPA
Path: utzoo!watmath!clyde!cbosgd!cbdkc1!desoto!packard!topaz!@RUTGERS.ARPA:ALDERSON@SU-SIERRA.ARPA
From: @RUTGERS.ARPA:ALDERSON@SU-SIERRA.ARPA
Newsgroups: net.works
Subject: Re: WORKS Digest   V5 #6
Message-ID: <553@topaz.ARPA>
Date: Thu, 7-Feb-85 18:39:01 EST
Article-I.D.: topaz.553
Posted: Thu Feb  7 18:39:01 1985
Date-Received: Sat, 9-Feb-85 07:45:25 EST
Sender: daemon@topaz.ARPA
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 13

From: Rich Alderson 

I have before me a copy of National Semiconductor's publication "The NS16000
Benchmarks."  They tested it out against a VAX 750 sometime prior to August,
1983 (the publication date).  It out-performed the VAX on an FFT, and had
comparable Whetstone times, when a 16081 (FPU slave processor) was attached.

Before guessing any further about your own set up, I would note that the 32082
MMU (formerly the 16082) has in hardware both Page Referenced and Page Modified
bits, allowing for LRU memory management.  The VAX lacks the Page Referenced
bit, and has to have same built into a table in the operating system.  This
slows it down badly.
-------