NJIT Planned Its First Computer Lab 65 Years Ago, Used for Dissertations
There were no digital computers at NJIT, known as Newark College of Engineering, in 1960.
This was not a unique situation. Most computers in 1960 were room-sized beasts performing logic through vacuum tubes. A few companies made smaller machines, when small was a relative term meaning something about the size of a Mini Cooper.
Large universities at that time might lease a mainframe, but most organizations including Newark College of Engineering instead purchased punch-card accounting systems for administrative use, along with mechanical terminals called teletypes to access computers located elsewhere.
Remote access in those situations was expensive. Professors and graduate students had priority. It was rare for undergraduates to get anywhere near a computer. The process of computing through a central processor located elsewhere was called timesharing, and its architecture should sound familiar compared to what we now call cloud computing.
But a sea change had been underway since the middle of the 1950s. Thanks to the transistor, which was cheaper, faster, more reliable, and vastly smaller than a vacuum tube, electronics of all kinds became easier to build and available to more people. The computer industry was no exception.
That opportunity was surely on faculty members' minds, when in 1954 — 65 years ago — a committee convened with one professor from each department to study how the college might establish a computer laboratory. The committee reported its conclusions to the Board of Trustees in 1960 along with comments on how owning a computer would help local industries, who in turn would help pay for it by making NJIT into an organization that could sell, rather than buy, any leftover computer time.
"It is recommended that a computer laboratory be established at the College. At the present time, the College is teaching several courses that are computer-oriented. In addition, time on computers is being purchased locally from industry and is being donated," board member Francis Keegan said, as recorded in NJIT university archives.
Keegan continued, "With the adoption of the new undergraduate curriculum, several additional courses utilizing computers will be taught. Several research projects conducted by the research foundation have need for this facility, and a request for a large block of computer time has been made by one of Rutgers Newark Division research directors. In addition, a survey of local industry shows considerable interest in various types of seminar courses in computer fields and in the purchase of unscheduled time periods."
The same year, Frederick W. Terman taught for one semester in the electrical engineering department. His father was Frederick E. Terman, better known as the legendary Father of Silicon Valley. The elder Terman was dean of the electrical engineering department at Stanford University and mentored his two most successful students — William Hewlett and David Packard — whose startup company in 1939 became the Hewlett-Packard we all know today. But the younger Terman, now 90 and retired in California, said he doesn't recall being involved in the discussion about building a computer center here.
One of the electrical engineering students at that time was Judea Pearl, who received his M.S. degree at NJIT in 1961. Pearl in the 1970s worked on artificial intelligence software at the University of California - Los Angeles. That work earned him a belated Turing Award in 2011, which is the academic computing field's highest honor.
The college finally acquired a computer in 1962. It was an IBM 1620. IBM's 1130, 1620, and 1800 systems were all popular at universities. According to IBM the 1620 could add or subtract 5-digit numbers 1,780 times per second. It could multiply up to 200 times a second and divide 56 times a second. Base models had 32 instructions. Programmers wrote most applications in Fortran, and the computer without any peripherals weighed 1,200 pounds.
The computer had 20,000 digits of decimal variable-word length memory. That's about 8 kilobytes of conventional binary digits based on the average size of real-world 1960s programming, said Carl Claunch, who recently helped restore a 1620 at the Computer History Museum in Mountain View, California. For comparison, 8 kilobytes is 131 thousandths of a gigabyte, or two-tenths of a second of the CeeLo Green song Forget You.
For comparison, 8 kilobytes is 131 thousandths of a gigabyte, or two-tenths of a second of the CeeLo Green song Forget You.
Only highly trained specialists would actually touch such a computer. Most users would submit stacks of punched cards, wait for their job to be processed, and hope for a bug-free result.
Donald Casseli was among the lucky few to operate the computer directly. "Following a four-year assignment in the U.S. Air Force, I entered Newark College of Engineering in September 1962. I immediately learned about their IBM 1620 Model I and was able to become part of a small group that learned machine language and worked directly with this computer," he wrote. "Years of study and other circumstances did not lead to a degree, but the education was used throughout my career."
The computer served many researchers throughout the 1960s. NJIT personnel gave lectures at a 1964 conference for the IBM 1620 Users Group. References to custom software written for the 1620 are found through dissertations archived on the NJIT website, such as The Rimo Filter, Prediction of Batch Heat Transfer Coefficients for Pseudoplastic Fluids in Agitated Vessels, and The Mechanism of Second Breakdown in Transistors.
It wasn't all business. In October 1965 students at male-dominated NJIT organized a computerized dating party with students from female-dominated New Jersey State Teachers College, which is now Kean University. NJIT students wrote the software and 1,500 people participated, an archived web page states.
The document continued, "The joint committee developed a personal profile that looks rudimentary in today's world, but nonetheless was effective in matching and according to a follow-up study several matches resulted in marriage."
Editor's note: this is part one of a three-part series about the history of computing at NCE/NJIT. Part two is here. Part three is here.