From lavelle@mbi.ucla.edu  Fri Feb 20 17:13:12 1998
Received: from theta1.ben2.ucla.edu  for lavelle@mbi.ucla.edu
	by www.ccl.net (8.8.3/950822.1) id RAA01477; Fri, 20 Feb 1998 17:00:45 -0500 (EST)
Received: from ewald.mbi.ucla.edu (ewald.mbi.ucla.edu [128.97.39.21])
	by theta1.ben2.ucla.edu (8.8.8/8.8.8) with SMTP id OAA17570;
	Fri, 20 Feb 1998 14:00:47 -0800
Received: from red5.mbi.ucla.edu by ewald.mbi.ucla.edu (5.65v3.2/1.1.10.5/09Oct97-1217PM)
	id AA16982; Fri, 20 Feb 1998 14:00:46 -0800
Message-Id: <3.0.32.19980220141654.0098ea00@mbi.ucla.edu>
X-Sender: lavelle@mbi.ucla.edu
X-Mailer: Windows Eudora Pro Version 3.0 (32)
Date: Fri, 20 Feb 1998 14:16:54 -0800
To: "Dale A. Braden" <genghis@darkwing.uoregon.edu>,
        cclpost <chemistry@www.ccl.net>
From: Laurence Lavelle <lavelle@mbi.ucla.edu>
Subject: Re: CCL:PC configuration for QM



At 12:22 PM 2/20/98 -0800, Dale A. Braden wrote:

>Dear CCL,
>
>I've looked at the CCL archives for info on this question, but the most
>up-to-date advice is important.
>
>I'm going to set up a PC/Linux system for doing quantum chemical
>calculations (probably using Q-Chem and Red Hat Linux 5.0), and the
>question is, What sort of hardware should I buy?  Rather than ask for
>answers to this broad question, I'd like to suggest a system and solicit
>criticisms of it. Here goes: 
>

>I plan on getting a dual-cpu PentiumII 333Mhz motherboard (these have the

>usual 512K L2 cache), 128Mb SDRAM, and two ultrawide SCSI drives (as large



After the CPU, cache is the most important. Accept nothing less than 1MB per CPU.

Digital did a demo here some months back and stated that CPU intensive jobs
show exponential dependence on cache size for their CPU's (don't know the
cutoffs).  They have/will release their latest 600MHZ alpha's with 4MB cache
modules per CPU as an option.

I would expect 1M cache/CPU for your high end Intel. I have heard of intel
motherboards taking up to 2MB cache/CPU. Bottom line is cache is very
important for CPU intensive work.


Laurence



<center>


""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""

Laurence Lavelle, Ph.D.

University of California Los Angeles

Molecular Biology Institute, and Department of Chemistry & Biochemistry

Laboratory of Structural Biology & Molecular Medicine

Los Angeles, CA 90095-1570, USA

 

Email:LAVELLE@MBI.UCLA.EDU

Phone (Lab): (310) 206-8270

Phone (Office): (310) 825-2083

Fax: (310) 267-1957

http://www.doe-mbi.ucla.edu/people/lavelle/lavelle.html

""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""


It will be a great day when schools have all the money they need,

and the military has bake day sales.




In nature's infinite book of secrecy

A little I can read.


</center>