At one of the most high profile auctions of a historic tech artifact yesterday (June 19, 2014), an early microcircuit failed to reach the minimum estimate of $1,000,000. Auction house Christie's pulled the device when bidding “only” reached $850,000. Even that is an extraordinarily high value number for such an arcane collectible that is difficult to see with the naked eye. What is the story behind it? Could this be the prodigal “lost” chip?
As early as 1952, British researcher G. W. A. Dummer stated that it was now possible to build miniature electronic circuits on single chips of semiconductor material without interconnecting wires. His own efforts failed but others succeeded in making working integrated circuits or ICs as they became known. In the mid 1950s, Rick Dill at IBM and Ian Ross and Arthur D'Asaro at Bell Labs built simple counters and in 1957 Torkel Wallmark fabricated a shift register at RCA. So why all the fuss over this later attempt that failed Dummer’s criteria on two counts? It used multiple chips and it employed interconnecting wires.
The earlier ICs used internal structures that limited their use to a restricted range of applications. Jack Kilby at Texas Instruments showed that semiconductor materials could be used to build more general purpose functions. He cut tiny bars of germanium containing transistors, diodes, resistors and capacitors, mounted them on glass slides and using fine gold wires connected them into three general purpose circuit functions - a phase shift oscillator, a flip-flop and a multivibrator. In September 1958, Kilby demonstrated his concept to his bosses at TI.
Even though TI could not make practical commercial products with this approach, Kilby’s determination and strong management support allowed the company to jump into the market for ICs as soon as others solved the remaining issues of putting it all on a single chip without wires. In 1960, the planar manufacturing process developed by Jean Hoerni at Fairchild Semiconductor in Silicon Valley enabled the fabrication of the first manufacturable general-purpose IC as patented by his boss Robert Noyce. This fundamental approach remains in use today. For his contribution, the Nobel committee awarded the 2000 Prize in Physics to Kilby. Hoerni and Noyce both died before the award.
Of the three Kilby circuits built in 1958, one was donated to the Chicago Museum of Science and Industry. TI loaned the phase shift oscillator to the National Museum of American History and, according to the Smithsonian Chip Collection, the “other was lost.” The Dallas Morning News noted today that the artifact offered at Christies was being sold by relatives of Tom Yeargan, the technician who hand-crafted it for Kilby. Could this be the prodigal lost circuit?
More on Silicon Valley History
Learn about more than 150 places in Silicon Valley associated with the origins of the digital and entrepreneurial revolution. Download the mobile app Silicon Valley Roots & Shoots from the iTunes App Store or Google play.