It's Not Easy Being Green (or "Red"): The IBM Stretch Project

History of Computing #2

Operator sitting at Stretch's ominously impressive operator console.

The Stretch CPU was over 33 feet long.

A typical Stretch installation. Along with the CDC 6600 from 1964, Stretch vies for contention as the world's first "super-computer."

Stretch Specs     Architecture: Aggressive uniprocessor parallelism Instruction Set: 735 instructions (incl. modes); variable field length Word Length: 64 + 8 check bits (SECDED) Memory: Magnetic Core (6 x 16KW, 2.1us cycle time); Disk: (2MW, 8Mbps); Magnetic Tape (12 x IBM 729 IV tape drives) I/O: 1,000 cpm (cards per minute) card reader; 600 lpm printer; 250 cpm card punch. Performance: approx. 500 KIPS (code dependent) Basic machine cycle: 300ns (3.3 MHz) Technology: Transistor (169,100) Size: Size: 2,500 square feet (size of average American home); CPU alone was 900 square feet (30 ft x 6 ft x 5 ft) Weight: Unknown; I estimate approx. 40,000 lbs. Number produced: 9 Cost: $7.78 million (1961 dollars) Power consumption: 21kW (processing units only)

"The fastest way to succeed is to double your failure rate." —Thomas J. Watson, Sr. Founder of IBM.

In the early to mid 1950s, IBM and UNIVAC, the only two large companies building computers, were beginning to consider transistors as a viable manufacturing technology for their products. Though the transistor effect had been discovered in 1947 at Bell Labs, vacuum tubes remained commonplace in computer hardware for several years. American manufacturers struggled to make a reliable, mass-producible transistor, and the transistors that were available were typically allocated to military needs.

Steve "Red" Dunwell and Werner Buchholz, two senior engineers at IBM, proposed a new machine, code-named "Datatron." Based on transistors, the machine would enable IBM to leap ahead of UNIVAC and would embody many new architectural concepts suggested by IBM's experience with its first series of machines: the 701, 702, and 704.

To the present-day observer, it may seem surprising that IBM was undergoing tremendous internal turmoil about its role in this new field. The public, however, had begun to associate computers with the UNIVAC name, much as we associate "Kleenex" with facial tissue. CBS's 1952 election coverage included a UNIVAC machine forecasting who would win that year's election. UNIVAC correctly predicted Eisenhower, but incredulous political analysts at CBS disagreed, so the computer's result was not made public until Eisenhower's victory was final. When former IBM customers started assigning key contracts to UNIVAC, IBM executives took notice.

100 Times Faster

In a famous memo to management dated October 25, 1954, Dunwell wrote: "The Datatron program is intended to assure IBM a preeminent position in the field of electronic data processing. To do this we must take a giant step and make substantial advances on all fronts." A team of senior IBM technical and management staff met to consider building what John von Neumann, then a consultant to IBM, had called (referring to a previous special-purpose computer, the IBM NORC) "the most advanced machine which is possible in the present state of the art." Besides allowing IBM to leapfrog its main competitor, Dunwell argued that this machine would allow IBM to rationalize its various computer lines—roughly divided along scientific and business lines—thus greatly reducing manufacturing costs and simplifying IBM's engineering and production processes.

After great internal debate and a contract from Los Alamos Scientific Laboratory, the project went ahead. Now code named "Stretch," the machine would be "100 times faster than the most advanced computer working today," and President Tom Watson, at the 1960 shareholder's meeting, proudly noted that the new machine could complete "100 billion computations in a day."

As news of the machine spread, several other customers placed orders.

The first machine (now officially named the IBM 7030) was delivered to Los Alamos on April 16, 1961. Although far short of being 100 times faster than competing machines, it was accepted and ran for the next ten years, with the then-astonishing average reliability of 17 hours before failure.

While customers were generally happy with the machine's performance, Stretch was considered a failure within IBM for not meeting its speed benchmark—with the consequence that IBM had to reduce the price from $13.5 million to $7.78 million, thus guaranteeing that every machine was built at a loss. Dunwell's star within IBM fell dramatically, and he was given fewer responsibilities—IBM's version of a gulag.

As time went on, however, attitudes within IBM changed about the lessons Stretch had to offer. From a lagging position in industry, IBM had moved into the forefront through the manufacturing, packaging, and architectural innovations Stretch had fostered. Dunwell's exile ended in 1966, when the contributions Stretch had made to the development of other IBM machines—including the monumentally successful System/360 product line—became evident. Dunwell was made an IBM Fellow that year, the company's highest honor.

A Successful Failure

Hundreds of IBM engineers had dramatically pushed the industry forward during the Stretch project, and Stretch alumni went on to contribute to some of the most important technologies still in use today. (One example is John Cocke, originator of the RISC architectural concept). Harwood Kolsky, designer of Stretch's lookahead unit, now emeritus professor at UC Santa Cruz, notes: "In the early 1950s the time was right for a giant step forward in computing. It was technology that pulled the computing field forward... This is where Stretch really stood out. It was an enormous building project that took technologies still wet from the research lab and forced them directly into the fastest computer of its day."

George Michael, a physicist and Stretch user at the Lawrence Livermore National Laboratory, notes that staff were very surprised that Stretch did not crash every twenty minutes. He calls the system "very reliable... it paid for itself in supporting the 1962 nuclear test series at Christmas Island."

The Stretch story is only one of many chapters in the history of computing demonstrating that our industry's triumphs are built upon the ashes of its "failures." Stretch is one of the hallmark machines—despite its being largely invisible to history—that defined the limits of the possible for later generations of computer architects. Looking at a list of Stretch milestones, you may recognize these many innovations in present-day products:

• Multiprogramming
• Memory protection

• Generalized interrupt system

• Pipelining

• Memory interleaving

• Speculative execution

• Lookahead (overlap of memory and arithmetic ops)

• Concept of a memory bus

• Coupling two computers to a single memory

• Large core memory (1MB)

• The eight-bit character (the "byte")

• Variable word length Standard I/O interface

Not heeding the lessons of history, microprocessor companies twenty or thirty years later "re-invented" most of these innovations. The Computer Museum History Center has parts of the original Stretch machine (serial number 1) from Los Alamos and a complete Stretch (minus core memory unit) from the Lawrence Livermore Laboratory.

Further reading

Bashe, Charles, et al. IBM's Early Computers, Cambridge: MIT Press, 1986, pp. 416-468.

Blaauw, Gerritt, & Brooks, Frederick, Computer Architecture: Concepts and Evolution, New York: Addison Wesley, 1997.

Buchholz, Werner, Planning a Computer System: Project Stretch, New York: McGraw-Hill Book Company, 1962. Out of print.

Dunwell, S. W., "Design Objectives for the IBM Stretch Computer," Proc. Eastern Joint Computer Conference, December 1956, pp. 20-22.

Retrocomputing, The IBM 7030