Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!ut-sally!im4u!rutgers!sri-spam!mordor!lll-tis!ames!necntc!necis!encore!fay
From: fay@encore.UUCP (Peter Fay)
Newsgroups: comp.arch
Subject: Re: Simulation eats MIPS
Message-ID: <1726@encore.UUCP>
Date: Thu, 2-Jul-87 16:17:23 EDT
Article-I.D.: encore.1726
Posted: Thu Jul  2 16:17:23 1987
Date-Received: Sat, 4-Jul-87 08:33:04 EDT
References: <40@b.gp.cs.cmu.edu>
Reply-To: fay@encore.UUCP (Peter Fay)
Organization: Encore Computer Corp., Wellesley Hills, MA
Lines: 25

In article <40@b.gp.cs.cmu.edu> jsp@b.gp.cs.cmu.edu (John Pieper) writes:
>... I am doing a relatively simple simulation...
>... Even though I can simulate a blazing 1 million clocks/second,
>the programs I am simulating run for 100's and thousands of mega-clocks.
>And I need to simulate 72 programs, each 10 times (varying a parameter).
>One program ran 2 giga-clocks before the simulation died and said "clock
>overflow". *That* took two days, and 2200 minutes of CPU time on an 8650.
>I don't know what I'd do if all I had was an 11/785. It would be real nice
>to have 720 Suns.

Try multiprocessing. CMU CS/RI has three multiprocessors, two of which are
Encore Multimaxes. These each have 14 32032's (could have 20 of them), and
will soon be upgraded to 323332's. They each then will be 28 MIP machines
with global shared memory. By the way, they are both running Mach
operating system.                                                    
                                                                       
                                                                       
                                                                       
			peter fay
			research
			encore computer
                                                         
                                  
                 
.