Nicholas Constantine Metropolis
Nicholas Constantine Metropolis, one of the pioneers in scientific computing, died in Los Alamos, New Mexico, on 17 October l999.
Born on 11 June 1915 in Chicago, Metropolis received both his BS (in 1936) and his PhD (in 1941) in chemical physics from the University of Chicago, where his thesis adviser was Robert S. Mulliken. Following graduation, Metropolis had a series of appointments during 1942 and 1943 that greatly shaped his future. First he was a research instructor with James Franck at the University of Chicago. He next became a member of the Manhattan Project under the supervision of Harold C. Urey. He then accepted a staff member position at the University of Chicago's Metallurgical Laboratory under Edward Teller. Teller persuaded Metropolis to become a theoretical physicist, thereby setting him on a path that would affect the world. Urey brought him to Los Alamos at the invitation of J. Robert Oppenheimer in early April 1943.
At Los Alamos during World War II, he endured, with Richard Feynman, the tedium of the slow, clanking, electromechanical devices used for hand calculations needed for weapons design. They actually hung out a shingle for their "repair business." Computing at the Los Alamos Laboratory during the war evolved from hand calculations, to punched-card computations, and then to electronic computing on the Electronic Numerical Integrator and Computer (ENIAC). The first electronic computer, ENIAC was developed at the University of Pennsylvania under the leadership of John Mauchly and Presper Eckert. John von Neumann's association with the ENIAC team, his knowledge of the Los Alamos effort, and his recognition of the profound implications of electronic computing led to his arranging for Stanley Frankel and Metropolis to run the first scientific problem--complex calculations involving the hydrogen bomb design--on the ENIAC in 1945.
The code that was to become the famous Monte Carlo method of calculation originated from a synthesis of insights that Metropolis brought to more general applications in collaboration with Stanislaw Ulam in 1949. A team headed by Metropolis, which included Anthony Turkevich from Chicago, carried out the first actual Monte Carlo calculations on the ENIAC in 1948. Metropolis attributes the germ of this statistical method to Enrico Fermi, who had used such ideas some 15 years earlier. The Monte Carlo method, with its seemingly limitless potential for development to all areas of science and economic activities, continues to find new applications. The Metropolis algorithm, first described in a 1953 paper by Metropolis, Arianna Rosenbluth, Marshall Rosenbluth, Augusta Teller, and Edward Teller, has recently been cited in Computing in Science and Engineering as among the top 10 algorithms having the "greatest influence on the development and practice of science and engineering in the 20th century."
Metropolis returned to Chicago at the end the war, but went back to Los Alamos in 1948 to take on the challenge of building a computer that would implement the rapidly developing concepts in digital computation. This effort resulted in the first (of three) Mathematical Numerical Integrator and Computer (MANIAC) at Los Alamos; MANIAC I became operational on 15 March 1952.
From 1953 to 1959 was an intense period of applications of the MANIAC and of the Monte Carlo technique to fundamental problems in physics, chemistry, biology, and mathematics. These efforts culminated in a series of seminal papers with distinguished coauthors in all those fields.
Metropolis returned to Chicago as founding director of the Institute for Computer Research, remaining there from 1957 to 1965. During this period, he was instrumental in developing imaginative new uses for computing. Perhaps his least known, though extremely important, achievement was the invention of online data processing in scientific experimentation. He designed and built--with a soldering iron in his hands--a computer that was coupled to the Navy cyclotron. This computer could receive and analyze data while an experiment was running, allowing the experimenters to modify their experiments during their allotted time. Metropolis also was active in organizing the data and storing the results on nuclear structure, which was a rapidly developing field of physics at the time. Eventually, however, his love for New Mexico brought him back to Los Alamos in 1965, where he remained the rest of his life, developing techniques and encouraging others to become interested in parallel computing.
A great architect of computational technique, Metropolis clearly foresaw the use of the computer as an experimental tool in mathematical discovery itself, as illustrated by his 1973 work with Paul Stein and Myron Stein on the universal nature of stable limit cycles associated with maps of the interval. The insight gained through this work would later help guide Mitchell Feigenbaum to the development of models of discrete dynamical systems.
Metropolis also was greatly concerned about problems with the foundations of mathematics itself that were revealed through computer use. He and the late Richard Hamming, friends from undergraduate days, spent hours discussing problems such as the concept of computable numbers. In the second half of his long career, Metropolis and the late Gian-Carlo Rota, renowned professor of mathematics and philosophy at MIT, became friends. This collaboration led to the publication of articles about foundational issues ranging from significance arithmetic, to new viewpoints on real numbers, to new insights into a subject as old as symmetric functions. In 1980, Los Alamos National Laboratory conferred on Metropolis the first senior fellow title. He was awarded the Pioneer Medal by the Institute of Electrical and Electronics Engineers in 1997.
Metropolis was an avid skier and tennis player until his mid-seventies. He also found amusement in creating original names for discoveries. For example, Emilio Segrč asked Metropolis--because of Metropolis's Greek heritage--to suggest names for the two elements Segrč and others had discovered. Metropolis did: "technetium" from the Greek technetos, meaning "artificial" for element 43, and "astatine" from the Greek astatos, meaning "unstable" for element 85. Metropolis also suggested "Monte Carlo" for the new statistical method of computation. And he created "MANIAC" for the new computer--hoping to end the acronym fad, as he put it--but, as he often lamented, with the opposite effect.
The Metropolis legacy will continue to flourish through the Metropolis Prize in Computational Physics, awarded yearly by the American Physical Society for the best dissertation in computational science. Also, a few years before his death, Metropolis established the Nicholas C. Metropolis Mathematics Foundation, whose purpose is to support and promote mathematics and computational science as exciting and challenging career choices, especially among younger students.
The international scientific community has lost one of its pioneers. His work in mathematics and computer science created much of the modern world large-scale simulation science. He will be sorely missed by his friends and colleagues at Los Alamos National Laboratory and elsewhere.
Nandor L. Balazs