Chapter 2 - Background
2.16 The Entrance of IBM - 1952
In 1950, it was far from obvious that computers would become an economic force capable of restructuring the very nature of how business is conducted or how entire economies work. In fact, it was not even clear that there was a need for computers in business. Many assumed their role would be limited to scientific uses. Business Week February 1963 reported prevailing opinion in 1950 was eight or ten of the giant “electronic brains” would satisfy everyone’s needs.19 Thomas J. Watson, President and CEO of IBM – the company that dominated the market for office machinery with approximately a 90% market share – should have known. Thomas J. Watson Jr. later writes: “My father initially thought the electronic computer would have no impact on the way IBM did business, because to him punch-card machines and giant computers belonged in totally separate realms.”20
After the public demonstration of ENIAC in 1944, the press reported its “nimble electrons made possible visions of computers helping to break the sound barrier, predict the weather, unlock the secrets of genetics, and design weapons even more terrifying than the Atomic Bomb.”21 On learning of ENIAC, the elder Watson called together his top engineers and told them that he wanted a better machine within eight months.22 He even conceded to use of vacuum tubes instead of the mechanical devices used in the Mark I. The result was the Selective Sequence Electronic Computer (SSEC), an electro-mechanical monster 120 feet long, with 12, 500 tubes and 21,400 mechanical relays. Not designed for commercial distribution, IBM dedicated it “to the use of science throughout the world.” It proved essential to the development of the atomic bombs used to end World War II. One housed prominently in IBM’s ground-floor, glass walled, Fifth Avenue, New York City offices could be used by all, without profit to IBM. Although it was the first calculator to run on software, IBM still suffered from a reputation “as a stodgy company that was wedded to punch cards and the past,” according to Watson Jr.23
On learning in September 1947 that Eckert-Mauchly had signed a contract to sell a UNIVAC to Prudential, Watson raged, ordering his top engineers to design a computer better than UNIVACin a week. When they proposed a design incorporating magnetic tape storage, Watson’s unyielding commitment to punch cards thwarted action, ending for now any effort by IBM to sell computers.24
In 1949, Watson Jr., recently promoted to executive vice president and frustrated by his father’s unwillingness to entertain any thought of responding to advances in electronics, and particularly computers, received help from an unsuspected quarter. Al Williams, a finance executive, reported on his study comparing research-and-development spending at RCA, General Electric, and other successful companies. On average, they spent three percent of revenue on R and D, while IBM only spent about two and a quarter percent. Reflexively, Watson commanded Watson Jr. to begin spending more money on research. In May of 1950, Watson Jr. promoted Wally McDowell to head an effort to build an electronics research operation with the order to hire “perhaps a few thousand”25 engineers and scientists.
But war, not management initiative, finally pushed IBM into the computer market. When the Korean War broke out, Watson “put the resources of IBM”26 at President Truman’s disposal. After studying the situation in Washington, James Birkenstock, manager of the Future Demands department, and scientist Cuthbert Hurd recommended to Watson Jr. that IBM build a “general-purpose scientific” computer. They felt they could find customers for thirty machines.27
This represented a break through in two regards: first, everyone besides Eckert-Mauchly were building one-of-a-kind machines; second, Birkenstock recommended that IBM fund the designs internally in order to own any potential patents.28 They estimated the design and prototype would cost three million dollars and the program three to four times that. The new computer – code named Defense Calculator, became the most expensive project in the company’s history, ten times the cost of SSEC.29
During 1951, while work progressed on the Defense Calculator, management feared a lawsuit by the Truman Administration’s “anti-big business” Justice Department – they did constitute approximately 90% of the highly visible punch card business. Their concerns were justified. On January 21, 1952, the Justice Department filed an antitrust suit against IBM claiming it constituted a monopoly. To Watson, whose prior experience at NCR with the Justice Department still left him bitter, there was absolutely no question but IBM would fight the unjustified claims. To his way of thinking, IBM represented only a very small percentage of all “numerical calculations done in business,” which seemed proof enough IBM did not constitute a monopoly. (This would not be the last monopoly suit the Justice Department would file against IBM.)
By early 1952, IBM had a working prototype of the Defense Calculator, but the production costs were significantly higher than projected. To cover costs, IBM had to double the monthly lease rate. To the “total amazement”30 of Watson Jr., customers paid the higher price. Clearly, demand for the product existed. “Customers wanted computers so badly that we could double the price and still not drive people away,”31 according to Watson Jr.
At the annual shareholders meeting in April 1952, IBM announced with great fanfare the release of an electronic machine, no longer named the Defense Calculator, but the IBM 701. Among the 150 top scientists and business leaders attending the ceremony was J. Robert Oppenheimer. He described the 701 as “a tribute to the mind’s high splendor.”32 In the promotional material, IBM claimed the new machine would “shatter the time barrier confronting technicians working on vital defense projects.”33
Yet before the IBM 701 came off the production line in December 1952, it earned another name, not of its choosing, but from the power of the new medium of television. On election night November 1952, CBS made public display of its use of a UNIVAC computer to predict election results.34 That night, the public made UNIVAC synonymous with computer. The 701 became known as IBM’s UNIVAC.
However, the 701 had an even bigger problem: it was inferior to UNIVAC. By the first 701 installation in the spring of 1953, UNIVAC had been shipping for 20 months. Even so, UNIVAC still proved more competitive than IBM management had hoped. The following table shows the 701 vastly superior in internal computing speeds, but since it used punch cards, and only punch cards, not magnetic tape as did UNIVAC, getting data into and out of the computer was significantly slower. (“ One-tenth the time on early computers was taken up in solving a problem while nine-tenths of the time was typically taken up by the card input-output and printing.”)35
In July 1953, under growing pressure to sell a commercial computer, IBM introduced the IBM 650 for delivery in December 1954. Designed by the Applied Science group, the 650 served those customers who “ did not want to make the big jump from a punched card system directly to a large-scale computer.”36 In being an easier computer to convert to and use, the 650 “changed IBM’s image from a producer of “IBM UNIVACs” to the leader in the industry.”37
Exhibit 2.21 TYPICAL COMPUTATION TIMES (IN MILLISECONDS)
|Function||UNIVAC I||IBM 701|
Then, in September 1953, IBM announced the commercial version of the 701, the 702, for delivery in 1955. But the 702 also suffered when compared to the UNIVAC for it used punch cards, not magnetic tape storage, and lacked adequate memory. “Our engineers and production mangers weren’t sure how to proceed,”39 remembers Watson Jr. One answer came from another government project underway at IBM since 1952, the SAGE project.
“Can IBM Keep Up the Pace?”Business Week, Feb. 2, 1963, p. 93
Thomas J. Watson Jr., “Father, Son & Co.,” Bantam Books, 1990, p.200
Watson Jr., pp. 200-201
Watson Jr., p. 203
Watson Jr., pp. 203-205
Ibid., p. 214
Ibid, p. 213
Ibid., p. 216
Ibid., p. 216
Ibid., pp. 216-217
Ibid., p. 243
The UNIVAC predicted Eisenhower over Stevenson by 438 electoral votes to 93, with a 100-1 odds of winning.Convinced the computer was wrong, CBS made the operators recalculate the prediction.When the new calculation predicted 8-7 odds in favor of Eisenhower, CBS reported that figure.The operators, however, realized they had made an error the second time, and that the initial prediction was correct.When Eisenhower won the electoral votes by 442 to 89, CBS had to admit the computer had predicted correctly and the UNIVAC earned its prominent place in computer history.
Ibid., p. 59
George Schussel, “IBM and REMRAND Part 2” Datamation, June 1965, p. 63
Ibid., p. 178
George Schussel, “IBM and REMRAND Part 2” Datamation, June 1965, p. 61
Watson, p. 259