Want to share an article with someone? Fill out the fields below and we’ll email them them the contents of the article you just read.
Please enter the word you see in the image below:
February 6, 2009
by Bruce Gilchrist
Editor’s note: James Pomerene, born June 22, 1920 in Yonkers, New York, died December 7, 2008 in Chappaqua. Bruce Gilchrist, a former resident of Chappaqua, knew him well and has written this recollection in his memory.
I will always remember Jim, who was one of my early mentors in the computer field. We were both in our thirties at the time; even then computers were a young person’s business.
At the end of World War II, John von Neumann had invited Jim to join a project at the Institute for Advanced Study in Princeton, which was designing one of the earliest stored program computers. When I arrived on the scene in 1952, the Institute for Advanced Study computer had just been completed under Jim’s leadership as chief engineer. Though when it comes to computers, as we have seen, nothing—then or now—is ever really completed.
In those days computers were complicated, unreliable monsters. Therefore, out of necessity, the 24-hour day was usually divided into two parts: the engineers had it for maintenance and improvement work during daylight and those of us working on applications had it at night. The only way bugs were detected and identified as hardware or programming related was to have good communications between the engineers and users. Fortunately we were few in number and all got along very well. I credit Jim with this as it was all too easy for us users to blame our analytical or programming mistakes on the computer.
Jim and I hit it off right away. One of our first joint projects was improving the performance of the notoriously unreliable Cathode Ray Tube memory, which had been designed by Jim based on work by engineers at Manchester University in England. Remember those were the days before magnetic cores existed, let alone chip memories. My contribution was to write a diagnostic program to identify errors during routine maintenance rather than have those errors spoil an applications run. Based on the error report generated, Jim and his associates were able to make the necessary adjustments.
Another joint project was to look for ways to speed up a computer’s arithmetic operation. Jim and another engineer, Sy Wong, contributed the electronic engineering know-how and I programmed various simulations. The end result was a joint paper entitled “Fast Carry Logic for Digital Computers” published in 1955. With typical modesty, Jim insisted that the authors be listed alphabetically, thus giving me more credit than I probably deserved.
In 1956 the Institute project ended and the staff dispersed. Jim went to IBM where he continued his very distinguished career and I moved to Syracuse University. It had been a great four years for me, made especially exciting due to the mentoring and friendship of Jim Pomerene.
Editor’s note: Subsequent to working with Gilchrist at Princeton, in 1956 Pomerene moved to IBM. He became an IBM Fellow in 1976. While at IBM, he received the IBM Outstanding Innovation Award in 1968; the IEEE Computer Society’s Pioneer Award in 1986; and the IEEE Edison Medal in 1993 “for outstanding contributions to the development of computer architecture, including pipelining, reliable main memory and memory hierarchies,” according to Bigelow Goldstein, a member of the National Academy of Engineering to which Pomerene was elected in 1988.
Bruce Gilchrist lived in Chappaqua from 1959 to 2008. He currently resides in Richmond, Virginia, with his wife, Bette.
Jim Pomerene, Bruce Gilchrist and Herman Goldstine taken at a reunion around 1970. Herman was the Assistant Director of the IAS project and later lived in Chappaqua for many years.
John Von Neumann with the IAS computer