1000-Pentium beowulf computer for genetic programming research (fwd)

Dan Brickley (Daniel.Brickley@bristol.ac.uk)
Wed, 11 Aug 1999 18:59:32 +0100 (BST)


---------- Forwarded message ----------
Date: Wed, 11 Aug 1999 13:50:33 -0300
From: John Koza <koza@smi.stanford.edu>
Reply-To: koza@stanford.edu
To: INDUCTIVE@LISTSERV.UNB.CA
Subject: 1000-Pentium beowulf computer for genetic programming research
Resent-Subject: 1000-Pentium beowulf computer for genetic programming research

Hello:

We have just posted photos of our recently installed 1,000-Pentium
Beowulf-style cluster computer system consisting of a server and 1,000
Pentium II 350-MHz processors. It is constructed entirely from "Commodity
Off The Shelf" (COTS) components.

This new machine is operated by Genetic Programming Inc., a privately funded
research group aimed at producing human-competitive results using genetic
programming. We are currently working in the areas of automated synthesis
of analog electrical circuits and controllers, problems in computational
molecular biology, various other problems involving cellular automata,
multi-agent systems, operations research, and other areas of design, and
using genetic programming as an automated "invention machine" for creating
new and useful patentable inventions.

There are now a number of instances where genetic programming has
automatically produced a computer program that is competitive with human
performance. Competitiveness with human performance can be established in a
variety of ways. For example, genetic programming may produce a result that
is slightly better, equal, or slightly worse than that produced by a
succession of human researchers working on an well-defined problem over a
period of years. Or, genetic programming may produce a result that is
equivalent to an invention that was patented in the past or that is
patentable today as a new invention. The fact that genetic programming can
evolve entities that are competitive with human-produced results suggests
that genetic programming may possibly be used as an "invention machine" to
create new and useful patentable inventions.

Each of the 1,000 processors of our parallel computer system has a Pentium
II 350-MHz processor. Each processor uses 64 megabytes of RAM (so that the
system as a whole has 64 gigabytes of RAM). The 1000 Pentium II processors
reside on 500 dual-CPU ATX motherboards and each motherboard is housed in a
standard mini-tower box. Each mini-tower box contains 128 megabytes of RAM,
a 100 megabit-per-second Ethernet NIC, and a standard 300W power switching
power supply. There is no hard disk, video monitor, keyboard, floppy disk
drive, or other input-output device associated with any of the 1,000
processors. The processors run the Linux operating system (Red Hat Linux
6.0). The communication between processors and between the server and the
processors is by means of 100 megabit-per-second Ethernet. Each group 40
processors (20 boxes) is connected to one 24-port 100 megabit-per-second
Ethernet hub. There are 25 hubs in the system. Each hub is connected on its
uplink to one of two 100 megabit-per-second 16-port Ethernet switches. The
two switches are connected to each other and to the server.

The server computer is also a dual Pentium II 350-MHz processor. The server
has 256 MB of RAM. It runs on the Linux operating system (Red Hat Linux 6.0
from VA Linux Research). The server contains a 14 GB hard disk, a video
display monitor, a floppy disk drive, a CD ROM drive, and a keyboard.

The system is booted using a DHCP message from the server to the 1,000
processors. We used the Beoboot software from Rembo Technology SaRL,
Geneva, Switzerland. The 500 boxes directly access the server's file system
using NFS.

Additional information about the machine and genetic programming (and job
opportunities) can be found at http://www.genetic-programming.com

John R. Koza

Consulting Professor
Stanford Medical Informatics
Department of Medicine
Medical School Office Building
Stanford University
Stanford, California 94305

Consulting Professor
Department of Electrical Engineering
School of Engineering
Stanford University

Phone: 650-941-0336
Fax: 650-941-9430
E-Mail: koza@stanford.edu
E-Mail: koza@genetic-programming.com

For information about GECCO-2000 (GP-2000) conference in Las Vegas on July
8 -12, 2000, visit:
http://www.genetic-algorithm.org/GECCO2000/gecco2000mainpage.htm