Lawson, Charles L. SIAM oral history

Charles L. Lawson discusses his professional life and career with the Jet Propulsion Laboratory, with particular reference to his contributions to mathematical software. He obtained an MSc in mathematics (under Peter Henrici) and a PhD (under Theodore Motzkin) from the University of California at Los Angeles. During his studies he worked at the Institute for Numerical Analysis at UCLA using the SWAC and then at NASA’s Jet Propulsion Laboratory (JPL). From 1960 to 1996, Lawson was an applied mathematician at JPL, specializing in mathematical software. Lawson discusses the work of his group in some detail, including the contributions of his colleagues Paul R. Peabody, Richard J. Hanson and Fred Krogh. He discusses the challenges involved in producing ephemerides needed for spacecraft missions and the state of celestial mechanics during the 1960s. He also recounts the development of computing at JPL, including the organization of computing facilities, the main machines in use, and the challenges posed by the arrival of minicomputers and workstations. Lawson was responsible mathematical library development within JPL, including the lab’s MATH77 Fortran library. Lawson, Hanson, Krogh, D.R. Kincaid and Jack Dongarra were responsible for the conception, design and implementation of the Basic Linear Algebra Subprograms (BLAS) which created a standard, machine-independent application interface to vector routines optimized for specific computer architectures. The BLAS made linear algebra software more efficient, portable and structured. BLAS development was conducted in conjunction with the LINPACK project undertaken by Argonne National Laboratory, and the two projects proved symbiotic. Outside JPL, Lawson was a well known member of the mathematical software community and discusses the development of the field as well as his personal contributions. He discusses Cleve Moler, Brian Ford and Gene Golub. In 1968 Lawson became one of seven authors of the book Computer Approximations, a standard work throughout the 1970s. Lawson took place in the early Mathematical Software conferences organized by John Rice during the 1970s and was an early member of the IFIP 2.5 Working Group on Mathematical Software. Lawson edited the ACM SIGNUM Newsletter from 1972 to 1976, and had previously edited the Scientific Applications Department of the Communications of the ACM.