Custom Search

Supporting Documents

A History of Computer Communications 1968-1988

Overview
Early Histories
Other Documents
Book

A History of Computer Communications 1968-1988

Computer History Museum

CHM Home Page
Oral History Archive

Newsletter
 

 

Early Histories
The Remarkable Growth in the Use of Computers

a.4 "Real-time" Computing -- The SAGE Project -- 1952 - 1958

 

The end of World War II did not bring to a close the impact of war making on the development of computer technologies. With the detonation of an atom bomb by the Soviet Union in 1949, the need to develop a continental air defense system to provide early warning and interception of enemy aircraft became an urgent priority. The branch of the military most responsible, the U. S. Air Force, turned to M.I.T. and the newly established Lincoln Laboratory at M.I.T for assistance. The project, code named SAGE (Semi-Automatic Ground Environment), encompassed every operation from radar sites in Northern Canada, to the system of transmitting the radar data to computers in the United States, and to the computers that would interpret the data, determine if an attack was underway and execute an appropriate response.

 

Once analyzing how enemy aircraft would be identified in the voluminous, and constantly changing, radar data, it became obvious any design would require "real-time" computational capabilities -- only at the time no computer existed, nor was it known how to design one, that would meet real-time requirements. Computers were simply too slow.

 

By the summer of 1952, the time had arrived to design the computer system. The Air Force authorized M.I.T. to solicit proposals with an objective to "manufacture, install and maintain several dozens of the systems -- systems of unprecedented complexity which employed heretofore unproven technologies." [40] The M.I.T. scientist responsible for the procurement, Jay Forrester, had led the development of the Whirlwind computer, the most advanced design of its time. (Whirlwind was the direct parent of the SAGE computer.) He held serious discussions with Remington Rand, IBM, RCA, Raytheon and Sylvania.

 

After lengthy debate, IBM decided to submit a bid and was awarded the contract to work with MIT Lincoln Labs to design the computer system for SAGE in October 1952. The following summer, any doubt as to the project's importance -- or urgency -- evaporated when the Soviet Union detonated its first hydrogen bomb. In September 1953, IBM, RAND Corporation, Burroughs and Western Electric were named prime contractors -- IBM to "to design, fabricate, support and maintain two prototype computers for the SAGE system." [41]

 

For IBM, SAGE represented the opportunity to learn state-of-the-art computer technologies from the most advanced computer development laboratory in the world. It also meant hiring hundreds, and soon thousands, of engineers and manufacturing personnel steeped in electronics and computers -- skills and experiences essential to its transformation from punch-card to computer manufacturer. But while IBM learned, staff at Lincoln Labs suffered. Lincoln Labs personnel, used to an environment rewarding individual initiative and creativity, now had to work within a bureaucracy, and a just assembling one at that. Norman Taylor, one of Forrester's most trusted managers remembers:  "IBM seemed awful stupid to us. They were still designing circuits like radio and TV circuits." [42]

 

SAGE impacted the fortunes of IBM almost immediately. The summer of 1954 IBM engineers were struggling with how to innovate the 702 into a competitive product. One seemingly insurmountable problem turned on memory. How to provide more of it? A problem faced by all computer designers. But Forrester had conceived the solution -- magnetic core memory. Upon learning of this seminal innovation, IBM announced two new computers in 1954, the 704 and 705, both with magnetic core memory. Acclaimed as a "creative masterpiece," [43] the 704 established IBM as the technological leader. Innovations included: core memory, tape storage, and the first commercial high-level programming language, FORTRAN.

 

By the summer of 1955, IBM completed the SAGE computer prototype. The AN/FSQ-7, or better known as the Q7, weighed 300 tons and occupied twenty thousand feet of floor space. [44] With  no guarantee of winning the production contract, Watson Jr. remembers: "I worked harder to win that contract than I worked for any other sale in my life." [45] He knew it meant government funding to finance mass production of computers within IBM. Pressing every advantage he had, Watson Jr. agreed to build a factory without a contract in hand, on only a handshake from Forrester. [46]

 

Even though SAGE was "celebrated as one of the great technical achievements of its day," [47] it was obsolete before fully operational in 1958. Because in August 1957, the Soviet Union launched an intercontinental missile that traveled the length of Siberia and, then, in October 1957, launched Sputnik I. The Soviet Union clearly had the technology to launch missiles armed with warheads into the United States from sites in Russia. But SAGE had never been designed to detect and intercept missiles. [48]

 

SAGE may not have been the success sought by the Air Force, but IBM scored a coup. Not only did the technology transfer and learning lead to the success of the 704, which clearly established IBM as the leader in the commercial market for computers, but SAGE pushed IBM to build factories and train thousands of workers before every one else. [49] IBM had leaped from Watson Sr.'s "stodgy company" to technological leader, largely on the back of government funding.

 

With the IBM 704 and the Q7 computers, the technology trajectory of computers had advanced significantly. SAGE innovations such as core memory, real-time response to multiple users, keyboard terminals, computer-to-computer communications, printed circuit board construction, and diagnostic and maintenance systems are standard features in all future computers. [50] Even so, computer technologies remained in their infancy, because they still relied on vacuum tubes.

 

Two other events in 1956 left permanent imprints on IBM. In January 1956, IBM signed a consent decree with the Justice Department settling the 1952 anti-trust suit. IBM agreed "to divest itself of a portion of its punch-card manufacturing equipment; sell as well as rent its office machines; refrain from tie-in sales agreements for cards and machines; and move its tabulating service bureau operations to a subsidiary (Service Bureau Corp.), barred from using IBM trademarks." [51] It also had to "license competitors under its patents at reasonable royalties." [52] By the time IBM complied with this settlement, punch cards had become an unimportant technology, product category and market-structure -- hardly worth four years of protracted negotiations with the justice department. But no one anticipated the transformative nature of computers.

 

A second event had more lasting impact. In May, Thomas Watson Jr. became CEO of IBM. He began an immediate effort to reorganize IBM -- which at the time resembled a pyramid with everyone reporting to the CEO. "It wasn't just that IBM had grown up under a single dominant founder. With the computer advance, its whole technological base was out of balance. It had to make the leap from electro-mechanical equipment -- basically, the slowly changing world of the machine shop -- to the galloping technology of electronics," [53] according to a Business Week article at the time.

 

Watson Jr., Albert L. Williams, president, and Louis H. LaMotte, chairman of the executive and finance committee redesigned IBM over the summer. Williams later recalls: "' We ended up with about 110 empty boxes in a new line-and-staff-structured management chart." [54] They picked 110 IBM executives to fill the boxes and invited them to a three-day retreat. Watson remembers: "We went in a monolith, and we emerged three days later as a modern, reasonably decentralized organization, with divisions with profit responsibility and clear lines of authority." [55]

 

By the end of 1957, IBM's organizational structure consisted of four computer divisions, which in 1958 were to have revenues of $700 million -- larger than all of IBM before re-organization -- and four smaller divisions. The four computer divisions sub-divided responsibilities into large computers; small computers and punch-card machines; R and D; and basically everything else: sales, installation and service. Product divisions were responsible for product innovation and manufacturing. The distinctions between large and small computers eventually became a major problem -- an artificial distinction introduced by IBM's two historical development trajectories.

 

Between 1956 and 1958, just as IBM asserted control of the commercial market for computers, vacuum tube technology, the basis of all First Generation computers, was about to become obsolescent as the logic component of computers. The first technological discontinuity of computers, the transistor, was about to happen.

 


[40]Nelson, p183-184 (71)

[41]Ibid., p. 184 (74)

[42]Glenn Rifkin and George Harrar, "The Ultimate Entrepreneur," Contemporary Books 1988, p. 22-23

[43]TBD

[44]Howard Rheingold, "Tools for Thought," Prentice Hall Press 1985, p. 142

[45]Watson Jr., p. 247-248

[46]Watson Jr., p. 247-248

[47]TBD

[48]SAGE went into operation and provide protection against bombers and aircraft. One outcome was a change in U.S. nuclear policy to assured mutual destruction.

[49]Ibid., p. 249

[50]Nelson, p. 185 (76)

[51]"IBM girds for battle," Business Week Jan 25, 1969, p. 36

[52]"Can IBM keep up the pace?," Business Week  February 2, 1963, p.93

[53]Ibid., p. 94

[54]Ibid

[55]Ibid.