The Fields Institute
Industrial Mathematics Seminar Series
Thursday, March 11, 1999
5:00 - 7:30 p.m.
Room 230, The Fields Institute
SCHEDULE
5:00 - 6:00 p.m.
Artificial Intelligence
"Deep Blue: IBM's Massively Parallel Chess Machine"
Gabriel M. Silberman, IBM Centre for Advanced Studies, IBM Toronto Laboratory
6:30 - 7:30 p.m.
"Some Lessons from Deep Blue"
Monty Newborn, McGill University
ABSTRACTS
"Deep Blue: IBM's Massively Parallel Chess Machine"
Gabriel M. Silberman, IBM Centre for Advanced Studies, IBM Toronto Laboratory
IBM's premiere chess system, based on an IBM RS/6000 SP scalable
parallel processor, made history by defeating world chess champion Garry Kasparov.
Deep Blue's chess prowess stems from its capacity to examine over 200 million
board positions per second, utilizing the computing resources of a 32-node
IBM RS/6000-SP, populated with 512 special purpose chess accelerators.
In this talk we describe some of the technology behind Deep
Blue, how chess knowledge was incorporated into its software, as well as the
attitude of the media and general public during the match.
"Some Lessons from Deep Blue"
Monty Newborn, McGill University
Deep Blue's triumph over Garry Kasparov was the culmination
of many years of work by a large number of computer scientists. The initial
work on chess programming began in the late 1950s, and over the next thirty
years, programs gradually improved. Beginning in 1989, IBM's Deep Blue team
synthesized this work and took the field to new heights with their SP system.
Along the road to defeating Kasparov, much has been learned on how to create
such sophisticated search programs, and this talk will focus on this and on
some of the surprises. One surprise, for example, is the triumph of brute-force
search over more selective search.
SPEAKERS
Gabriel M. Silberman is Program Director for
the Centre for Advanced Studies (CAS) at the IBM Toronto Laboratory. Dr. Silberman
comes to CAS from the IBM T.J. Watson Research Center, Yorktown Heights, NY,
where he held various research positions from 1990 to 1997.
While at the T.J. Watson Research Center, Dr. Silberman was
closely involved in the Kasparov vs. Deep Blue matches, both in 1996 and for
the rematch in 1997. His group provided the computing and audio-visual systems
support infrastructure, both on- and off-site. In 1997, Dr.Silberman served
as Deep Blue team co-ordinator and liaison to the Kasparov team. Dr. Silberman
research activities at T.J. Watson also included work on Very Long Instruction
Word architectures and compilers, as well as general computer architecture
and systems issues.
Monty Newborn is a Professor of computer science
at McGill University where he teaches courses on AI, automated theorem proving,
and parallel search. He served as chairman of the ACM Computer Chess Committee
since its inception in the early 1980s. That committee organized the first
Kasparov versus Deep Blue match (The ACM Chess Challenge) and officiated at
the IBM Kasparov Versus Deep Blue Rematch. Newborn is the author of Kasparov
Versus Deep Blue: Computer Chess Comes of Age (Springer-Verlag, 1997), and
four other books on the subject of computer chess. His program, The Great
Theorem Prover, competed in this year's automated theorem-proving contest
in Lindau, Germany.
OTHER INFORMATION
The Industrial Mathematics Seminar is offered to any interested participant
-- no reservation is necessary.
