This tiny molecular computer made up of 47 modules containing 587 semiconductor networks was dwarfed by conventional transistorized computers that performed identical electronic functions.

First IC-based computer developed for US Air Force

After Jack Kilby invented the integrated circuit in 1958, TI President Pat Haggerty immediately began searching for customers who might have a product or system in need of the microminiaturized devices. In 1959, the integrated circuit was little more than a laboratory creation. Without contracts, TI would not have the funding needed to develop and expand the program, even though integrated circuits potentially could reduce the space and power used in electronic equipment and improve reliability.

Kilby and Charles Phipps, a key manager in TI’s strategic planning and marketing of integrated circuits, developed a good working relationship with Richard D. Alberts, chief of the transistor section at the Air Force Electronic Components Laboratory in Dayton, Ohio. Phipps described Alberts as “somewhat of a maverick,” but someone who “did have the technical foresight and imagination to see where he was going, and who relished twisting the tails of the other services and some of the people above him in the Air Force.” Alberts was intrigued by Kilby’s concept of silicon integrated circuits, and he defied established Air Force priorities by funding a small TI research program.

Previously, Kilby had faced the task of making each integrated circuit by hand. But Alberts circumvented Air Force policies a second time, providing additional funding for TI to research manufacturing processes. Willis Adcock, whose research and development lab produced the first integrated circuits, pointed out,
“I think we would have dropped the program had it not been for the Air Force’s support.” Alberts made it possible for TI to develop a pilot manufacturing line, which made a handful of devices that were ultimately sold for $450 apiece.

Texas Instruments had not been able to convince skeptics that integrated circuits were reliable. They thought that if you multiplied the yield of each component in an integrated circuits together, the overall yield would be so low you could never expect to build a good unit. Also, some feared the amount of heat dissipated within a densely packed group of circuits would keep them from functioning properly. Alberts and Phipps decided TI needed to build a model to demonstrate the viability and reliability of the technology. Alberts backed the project by funding $600,000 — enough to let TI build a small digital computer using integrated circuits. It was a “bootleg” project that Alberts decided to keep secret. The Air Force could not interfere with a project it didn’t know existed.

Harvey Cragon, an engineer in TI’s defense business, was assigned the task of building a tiny 10-bit computer capable of assorted mathematical functions. Cragon later said, “Along about that time, I hired Joe Watson. The two of us sort of designed the thing, and Jack built the circuits.” To refute critics of integrated circuits who said they would never be practical because of yield and power dissipation, the demo computer was built as small and dense as possible. A stacking concept was used to interconnect five or six integrated circuits into each module. This packaging scheme was years ahead of its time — a decade later memory stacks were used in the industry to save space in electronic equipment. And, the little computer was the first to employ solid state memory.

When completed, the computer used 587 integrated circuits and weighed a mere 10 ounces. To focus attention on its size, TI built a second unit offering the same proficiency, but assembled it with discrete solid-state devices. This machine was 150 times larger than the integrated circuit computer and almost 50 times heavier. And it used 8,500 discrete components, 14 times more than the unit built with integrated circuits.

The demo and tour of the tiny computer helped tip the scales in TI’s favor and made it possible to start applying integrated circuits to defense equipment in 1961 and to Minuteman and other programs in 1962.

 
The first molecular computer using 587 solid circuits
  Print Frame Bookmark

© Copyright 1995-2007 Texas Instruments Incorporated. All rights reserved.
Trademarks
 |  Privacy Policy  |  Terms of Use