Roger Hilsen Koch, a research scientist and manager at the IBM T.J. Watson Research Center in Yorktown Heights, NY, died of a heart attack on August 4, 2007 while riding his bicycle. He was an outstanding scientist with a command of computers, programming, and measurement and analysis techniques, coupled with a deep understanding of theoretical principles.
Roger was born on November 1, 1950 in Saginaw, Michigan. He displayed an early interest in science and engineering. When his small high school in Edmore, Michigan did not offer chemistry, he taught himself the subject. Roger's ability to self-learn and his insistence on deeply understanding all problems that he tackled were characteristic of his entire career. Roger earned a B.S. from MIT in architecture (1972). After graduating, he was offered a job tracing aspects of the history of technology at the Sturbridge Village Museum in Massachusetts. Rather than study history, however, he chose to contribute to technological developments, and took a design job at the Van Dyck Corporation in Southport, Connecticut. Throughout his career, Roger retained interests in the history of technology and in the elements of good design.
Subsequently, Roger decided to move into physics. In 1977 on an NSF fellowship, he entered graduate school at the University of California, Berkeley, where he worked with John Clarke. Together with Dale Van Harlingen, Roger measured quantum noise in Josephson tunnel junctions and, based on his results, carried out simulations showing that SQUIDs (Superconducting Quantum Interference Devices) should be quantum limited amplifiers. He devised a technique still widely used for reducing the infamous 1/f noise in SQUIDs. Roger would return many times to the study of noise in various physical systems throughout his career.
Roger joined IBM in 1982, immediately after receiving his Ph.D., and remained at the Watson Research Center for the rest of his career. Roger was one of the first to develop a general purpose data acquisition and analysis program utilizing the personal computer. Although only an off-shoot of his experimental research, the program was used at various IBM sites for many years. Long before it became commonplace, Roger ran complex experiments and massive simulations from home, on the road and during dull moments at IBM meetings.
In addition to his research on low transition temperature (Tc) Josephson junctions and SQUIDs, Roger worked on an astonishing number of other topics. Among his most widely cited papers are his observations of vortex-glass superconductivity in high-temperature superconducting films and the first measurements of high critical currents in these materials, dynamics of nanomagnetic systems, including spin-transfer induced switching, and STM studies of point defects in silicon. He wrote many pioneering papers on high-Tc SQUIDs. He was well-known for his studies of noise in a variety of physical systems, including SQUIDS, spin-glasses, metallic films near the percolation threshold and metallic systems undergoing electromigration. His work with SQUIDS led him to consider the problems of magnetic field detectors in general, and he made and characterized improved flux-gate magnetometers. He also co-authored papers on chaos, geophysics, resonant tunneling, scanning tunneling microscopy, neural networks, and submicron structures.
While an outstanding physicist, Roger also enjoyed engineering, and contributed his time and talents to developing novel measurement techniques to address practical problems of manufacturing process control. Most notably, he connected features of 1/f noise to incipient electromigration failures.
In recent years, Roger led an effort to develop essential elements of a quantum computer based on his unique approach using flux-controlled Josephson tunnel junctions. This project now involves a significant team at IBM Research. Although his contributions to quantum computing have been cut short, his invention of a control scheme for Josephson junction qubits based on a superconducting resonator is already widely cited, as is his theoretical treatment, with David DiVincenzo and Guido Burkard, of multilevel dissipative dynamics of circuits composed of superconducting qubits. The latter is an essential guide to the design of quantum coherent circuits.
Roger was respected as a superb and creative experimentalist who could hold his own with the best theorists in the field. He had great personal integrity and a disarming directness, and inspired many others through his dedication to science, his creativity, and his warm personal style. His many friends and colleagues will sorely miss him.
Roger was a devoted husband and father. Spending time with his wife, Inga, and his young son, Alexander, was the one thing that pulled him away from the laboratory. In addition to Inga and Alexander, he is survived by his brother, Jonathan, and his parents, Arthur and Loretta Koch.
