John Cocke | Biography

John Cocke laid the groundwork for today's generation of super-computers with his work in reduced instruction set computer technology, or RISC. As a researcher for International Business Machines (IBM), Cocke pioneered the theory and applications for software that translate user programs into machine-readable systems, or compilers, and simplify instructions in order to reduce operating cycles.

The youngest of three sons, Cocke was born May 25, 1925, in Charlotte, North Carolina. As a youngster, he was a poor student, and his parents hired a private tutor to instruct him while he was in elementary school. Cocke's lack of enthusiasm and curiosity, however, was confined to the classroom, and he tinkered with mechanical solutions to problems. Although he failed in his attempt to develop a gadget to catch birds by spraying salt on their tails, he jerry-rigged an electromagnetic window-washer and a hydraulic pipe wrench. In high school, he discovered science and mathematics and, perceiving their potential as tools for problem-solving, he developed abilities in these fields to match his interest in finding better, faster ways to tackle tasks.

Cocke entered Duke University in Durham, North Carolina, under the auspices of the U.S. Navy's V-12 program. The V-12 program was instituted during World War II in an effort to meet the military's need for college-educated officers. Cooke graduated with a bachelor's degree in mechanical engineering in 1946; he worked for several companies--including General Electric Company--until 1952, when the Navy called him to service. Cocke served two years with the Navy and returned to Duke in 1954, where he completed a doctorate in mathematics in 1956.

Cocke joined IBM in 1956, when computer science was in its infancy. The field was so new there were few, if any, rules governing research, theory, applications, and procedures. The freedom Cocke and colleagues such as Frederick P. Brooks, Jr., and Harwood Kolsky enjoyed spawned a productive period of scientific discovery. Cocke and Kolsky developed one of the first simulation programs used in computer design, and Cocke helped craft a "look-ahead" feature for programming.

Assigned to IBM's STRETCH project team, Cocke combined academic training with natural intuition to tease out solutions to computing problems. Many innovations envisioned and developed by Cocke and his colleagues are standard on today's high-performance computers and have played a key role in the development of super-computers.

By the mid-1960s, Cocke was working on advanced computer system design. He demonstrated how multiple instructions could be executed in a single cycle and developed a new branching structure that eliminated bottlenecks which previously hampered the performance of look-ahead systems. He also developed compiler techniques for look-ahead machines that dramatically improved the code produced-techniques that served as the foundation of optimization strategies essential for today's systems. Cocke contributed to advances in logic simulation, parallel processing or "pipelining," and large-system architecture. As an academic field supported by a legitimate research base, computer science grew rapidly, and as the field grew, researchers turned their attention to creating systems that were faster and physically smaller.

In the 1970s, Cocke led a team charged with developing a small computer based on RISC architecture. This search was fueled by practicality; in the 1960s, computer systems with the memory and power of modern personal computers were physical behemoths that took up entire rooms (sometimes whole floors) in office buildings. Size was a sticking point for sales; developing a computer that could do more tasks in less space was an industry imperative.

Cocke and his team at IBM's T. J. Watson Research Center in Yorktown Heights, New York, tackled the challenge. In the late 1970s, the team introduced the 801, a computer system that integrated RISC and optimizing compiler technology in a (for the time) modestly-sized package that came close to executing an instruction per cycle. The 801 was a prototype that established IBM as a market leader and served as the basis for applications that made modern high-performance systems and workstations possible.

Cooke managed to find time for other achievements. He and Richard Malm developed a special-purpose, high-speed logic simulator; descendants of that simulator are used today to verify and fix logical design before it is imbedded in silicon. Cocke also probed signal processing, cryptography, magnetic storage devices and coding processes, switching systems, and the mathematics of speech recognition.

The 1970s were also Cocke's most productive for publishing. In addition to numerous articles and papers, Cocke and his colleague Jack Schwartz published Computers, a series of IBM programming books that were widely distributed and proved to be highly influential beyond the academic community.

Cocke's work earned him numerous honors. IBM recognized his contributions in 1972 by naming him an IBM fellow, the company's highest honor for scientific and technical achievement. In 1976, he received the ACM Programming Systems and Language Awards; he was elected to the National Academy of Engineering in 1979 and named a fellow of the American Academy of Arts and Sciences in 1988. The recipient of the ACM/IEEE Computer Society Eckert-Mauchly Award in 1985, the National Medal of Technology in 1991, the National Medal of Science in 1994, and the IEEE John von Neumann Medal in 1994, Cocke was named inventor of the year by the Intellectual Property Owners Association in 1992.

When Cocke retired in 1993 after a 37-year career at IBM, he owned 22 patents.

Cocke died after a long illness in Valhalla, New York, on July 16, 2002. He was 77.