The Man Behind the Microchip by Leslie Berlin

Author:Leslie Berlin [Berlin, Leslie]
Language: eng
Format: epub
ISBN: 9780195163438
Publisher: Oxford University Press
Published: 2005-09-15T04:00:00+00:00

9

The Edge of What’s Barely Possible

Andy Grove spent much of 1972 worrying. In February he wrote, “We are not doing well at all! Momentum broken, refresh [problem with the 1103] still with us, bipolar yield still (or again) nowhere!” He kept detailed tabs on Intel’s competition, noting that the 4,000 1103s ordered by Texas Instruments were not going to their computer group, and announcing how many semiconductor memories had been shipped by TI-spinout Mostek in a given quarter. In December, he was concerned because the comments offered at the executive meeting to review the operations plan he laid out for 1973 were “brief, between long periods of silence and sighs.”1

Noyce, on the other hand, took a wider view and liked what he saw for most of 1972. Intel was building about 100,000 1103s each year and still could not keep up with demand. By year’s end, the 1103 was the best-selling memory chip in the world. It was so terribly difficult to reverse engineer that Intel had almost no competitors, outside of its official second source MIL, for more than a year. When Intel increased the size of its wafers and MIL tried to follow suit, the Canadian company lost the process entirely. This left Intel with a near-monopoly on the 1103. Effectively all of Intel’s $23.4 million revenue in 1972 came from sales of this one device.2

A tentative foray into building complete memory systems for customers was proving very profitable. The microprocessor had brought Intel hundreds of new customers, most of them small firms with small orders, but the trend nonetheless augured well for the future. Noyce expected that once Intel could build a more powerful microprocessor, the market would take off. And Intel had that processor under development. The company was building new facilities on the Santa Clara campus, a fab in Livermore, California, and an assembly operation in Penang, Malaysia. Noyce was considering adding an on-site day care center to the Santa Clara campus as a way to attract the largely female workforce for the fabs, but a series of informal discussions he held with various Intel employees never really went anywhere.3

In July, Intel acquired a six-man digital watch company called Microma for roughly $2 million. Noyce and Moore predicted that in the next five years, some 200 million digital watches would be sold around the world. Every one of them would require elements in which either Microma or Intel was expert: liquid crystal displays, detailed assembly, and a type of MOS chip called CMOS. Intel, true to its unwritten policies, was going to try to capture this market before it attracted attention from anyone else. Noyce expected Intel would, in the next few years, sell some $20 million in watch modules, 20 percent of which would go into Microma watches. The rest would be sold to other watch manufacturers.4

In November, Noyce invited one of the Grinnell trustees, Sam Rosenthal, to tour the new fab and independently assess the merits of the college’s investment. “Boy Noyce showed

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.