Successive over-relaxation

Lore & Background

The method of successive over-relaxation was devised simultaneously by David M. Young Jr. and Stanley P. Their work aimed at automatically solving linear systems on digital computers, a task that required a method suitable for programming. Over-relaxation methods had been used before, such as those by Lewis Fry Richardson and R. V. Southwell, but these earlier methods were designed for human calculators and required expertise to ensure convergence, making them inapplicable for digital computers. These aspects are discussed in the thesis of David M. Young Jr.

Reader's Guide

Successive over-relaxation (SOR) is a significant development in numerical linear algebra because it provided a way to accelerate the convergence of iterative methods for solving linear systems, specifically as a variant of the Gauss–Seidel method. Its importance lies in its automatic applicability to digital computers, unlike earlier over-relaxation methods that relied on human judgment. The method remains a standard technique in scientific computing, particularly for large sparse systems. Its legacy is tied to the broader shift toward algorithmic methods that could be programmed, enabling the efficient solution of linear equations in fields such as engineering and physics.

Did You Know?

• SOR is a variant of the Gauss–Seidel method for solving linear systems.
• Earlier over-relaxation methods by Lewis Fry Richardson and R. V. Southwell were designed for human calculators.
• The method was developed for automatically solving linear systems on digital computers.

Prodigious Beginnings in Budapest

His father, a banker with a law doctorate, had moved the household from Pécs to the capital in the late 1880s, where they occupied an eighteen-room apartment above the Kann-Heller offices. As the eldest of three brothers, young János was tutored at home in English, French, German, and Italian. Family legend says that by age eight he already knew differential and integral calculus, and by twelve he had read Borel's treatise on functions. His father insisted he attend school at his own grade level but permitted private tutors. At fifteen he began advanced calculus with Gábor Szegő, whose wife recalled the analyst coming home in tears, stunned by the teenager's speed. By nineteen, von Neumann had published two major papers, the second giving the modern definition of ordinal numbers that superseded Cantor's. He concluded his gymnasium years by winning the national Eötvös Prize in mathematics.

Dual Degrees and the Göttingen Circle

When it came time to choose a career, von Neumann's father wanted him in industry rather than pure mathematics. He asked his friend Theodore von Kármán to talk John out of a mathematical path, and the two settled on chemical engineering. Simultaneously, he enrolled as a Ph.D. candidate in mathematics at Pázmány Péter University in Budapest, where his thesis produced an axiomatization of Cantor's set theory. A Rockefeller Foundation grant then carried him to the University of Göttingen to study under David Hilbert.

Wartime Service and the Nuclear Defense Establishment

During the Second World War, von Neumann joined the Manhattan Project, where he developed the mathematical models underlying the explosive lenses critical to the implosion-type nuclear weapon. His consulting work extended well beyond the laboratory: before and after the war he served the Office of Scientific Research and Development, the Army's Ballistic Research Laboratory, the Armed Forces Special Weapons Project, and Oak Ridge National Laboratory. In the 1950s, at the height of his influence, he chaired several Defense Department committees, including the Strategic Missile Evaluation Committee and the ICBM Scientific Advisory Committee. He also sat on the Atomic Energy Commission, the body overseeing all atomic energy development in the United States. Working alongside Bernard Schriever and Trevor Gardner, he played a key role in designing and developing the country's first intercontinental ballistic missile programs. At that time he was regarded as the nation's foremost expert on nuclear weaponry and the leading defense scientist at the Department of Defense.

A Mind Without Boundaries and an Enduring Mark

Few mathematicians of the twentieth century matched von Neumann's range. He built the mathematical framework of quantum physics, advanced functional analysis, and introduced or codified game theory. His work on cellular automata, the universal constructor, and the architecture of the digital computer anticipated entire fields of computation. Notably, his analysis of the structure of self-replication preceded the discovery of DNA's structure. Colleagues across physics, mathematics, and the broader sciences praised both his contributions and his sheer intellectual ability. The breadth of his impact was recognized with the Medal of Freedom, and a crater on the Moon now bears his name. Whether in Budapest's reading rooms, the rain-wet streets of Göttingen, or the corridors of the Pentagon, von Neumann left a mark that spanned pure abstraction and the most consequential applied sciences of the age.