The impetus for the Defense Calculator came from the Korean War. The Korean conflict, begun in 1950, lent urgency to the push for new planes and weapons that would operate at higher speeds, higher temperatures, and with greater precision. Designing and producing them meant another surge in demand for engineering calculations, only five years after the end of World War II. The Pentagon and its corporate suppliers were sophisticated customers with deep pockets, but they were few. And it was not yet clear that there would be a lucrative market for the tireless calculating capacity of electronic computers outside the defense establishment.
IBM DividedWithin IBM, there were two disparate schools of thought. The enthusiasts, led by Thomas Watson Jr., the scientists, and younger managers, understood that the demand for computing would spread broadly, and Remington Rand's UNIVAC computer was showing the way, having sold a machine to the Census Bureau. The skeptics at IBM included the chairman Thomas Watson Sr., and much of the senior management. Customers would be scarce, they worried, and manufacturing such a technically challenging machine would drain the companys engineering resources. The plan for the new machine was approved in early 1951, but in deference to the in-house skeptics the machine was called the Defense Calculator, suggesting that it was a special project in support of the war effort.
IBM 701Despite the name, the Defense Calculator was a stored-program computer, and so it was a general-purpose machine, awaiting only programming instructions to tackle all manner of problems. Indeed, by the time the machine was unveiled to the public in April 1953 it had undergone a name change, becoming the IBM 701, the first of the 700 series which firmly set IBM on its way to being the world's dominant computer maker. The 701 was compact and stylish by the computing standards of the time. The system was a collection of stand-alone units that looked something like a department-store display of 1950s-vintage kitchen appliances, the pair of tape readers resembling big cabinet-style televisions; the printer, an oven; the cathode-ray storage unit, a refrigerator. Yet it was the speed of the 701 for its day that most impressed the private audience who gathered in Poughkeepsie in the summer of 1952. They brought with them sample programs, encoded and punched onto paper tape. They each got a shot at the computer, recalled Cuthbert Hurd, an IBM executive who was there. They would feed a program into the computer and, bam, you got the result... We all sat there and said, How are we going to keep this machine busy? It's so tremendously fast. How are we going to do that?
The mammoth, costly IBM machines of the 1950s, to be sure, possessed a tiny, tiny fraction of the computing firepower of even a handheld computer today. But the 701 was a speed demon in 1952, so IBM found itself facing the digital paradox the total interdependence of two very different disciplines, computer software and computer hardware, the yin and the yang of computing. The answer to Hurd's question about how to keep the fast computer busy was simple enough: put more problems on the machine. But there was a bottleneck, and it was programming.
Preparing an engineering or scientific problem so that it could be placed on a computer was an arduous and arcane task that could take weeks and required special skills. Only a small group of people had the mysterious knowledge of how to speak to the machine, as if high priests in a primitive society. Yet there were some heretics in the priesthood, and one of them was a young programmer named John Backus. Frustrated by his experience of hand-to-hand combat with the machine, Backus was eager to speed things up and somehow simplify programming. I figured there had to be a better way, he recalled nearly five decades later at his San Francisco home, which overlooks the Golden Gate Bridge. You simply had to make it easier for people to program.