Milestones That Mattered: The planar IC—revolution underestimated
By Maury Wright, Editor in Chief -- EDN, April 27, 2006
Born in the labs of Fairchild Semiconductor and Texas Instruments in the 1958 time frame, the planar IC finally emerged in 1960; Fairchild shipped it in March 1961. Robert Noyce, the co-founder of Fairchild, and Jack Kilby, an engineer at Texas Instruments, both built early ICs in the lab, and both received patents, which the two companies subsequently cross-licensed.
 From EDN, October 1, 1960: Micrologic elements being developed MOUNTAIN VIEW, CALIF.—High-speed, low-power digital computer logic building blocks are under development at Fairchild Semiconductor Corp. To be available early next year, the family of solid-state micrologic elements will handle all the logic-function requirements of a digital machine, no other components being required. The micrologic elements are made by diffusing planar transistors and resistors into a solid continuum of silicon; element intraconnections are then deposited on the surface. The low cost resulting from batch-processing will mean lower first cost of a computer logic section. It is expected that reliability of the micrologic elements will be at least as good as that of a well-engineered contemporary logic circuit performing the same function. Click here for the complete text of the 1960 article. |
Early recognition of the significance of the IC was underwhelming. But the planar IC became perhaps the single innovation that has had the greatest impact on the high-technology industry and arguably our entire society because electronics underlie advancements in everything from medical procedures to transportation.
EDN's early coverage of the IC explicitly questioned its value. In "Micrologic elements being developed," the cover story from our Oct 1, 1960, issue (see the excerpt on the right or read the full text), our editors downplayed the advantages of size that the IC offered. The staff noted that other system elements, such as power sources, dwarfed the size of electronics modules. Therefore, they believed ICs were unnecessary for hot applications of the day, such as spacecraft.
Even Kilby has made it clear that the implications of the IC were underestimated. Years later, he said, "What we didn't realize then was that the integrated circuit would reduce the cost of electronic functions by a factor of a million to one. Nothing had ever done that for anything before."
Fairchild's marketers at the time had clearly zeroed in on computers as the primary application for their "micrologic elements." Based on most historical accounts, however, few realized the impact that these predecessors of the 74xxx family of standard-logic ICs would have in simple combinational- and sequential-logic applications. Still, Fairchild was clearly right in realizing that ICs would be key enablers of computers.
Noyce went on in 1968 to co-found Intel, where the microprocessor was born and nurtured. Noyce oversaw the microprocessor-development project. Even before Intel's birth, Gordon Moore, another co-founder, had stated in 1965 what would become known as Moore's Law. Today, Intel and others continue on the Moore's Law path, and Intel even promises to outpace the law performancewise using dual-processor chips. At a recent meeting with several editors from EDN, Electronic News, and Electronic Business, Stephen Smith, vice president and director of desktop-platform operations at Intel, stated, "IC cost has gone from $5000 to 1 billionth of a cent per transistor over 50 years."
-
If you look at the timetable outlined in that excellent historical perspective, you will find a clear cause for the evidence of truth to Moore's Law and the advent of the dominance of planar ICs.
These devices and Moore's postulate required an adequate, challenging proving ground in which to grow these technologies (or to micronize, as the case actually was). The United States' space program, with an immediate requirement for advanced data processing and power management, was that proving ground.
The global pursuit of the technological heritage of that time period has driven almost every advance for the last 40+ years. Only the total commitment of resources by a nation or group of nations can recreate the level of advances that were experienced during the 1960s-1980s.
It is time for a new challenge.
Whichever nation or nations select the most effective technological proving ground today - whether that is putting men on other planets or unlocking the components and mechanics of subatomic particles - will be the ones that set the course of development for the next half-century. And probably beyond.
And the next generation of companies, like Fairchild in 1960, will discover the market for such advances likely to exist in places they'd never imagined.
Rob Purdie - 2006-3-5 08:52:00 PDT -
Milestones That Mattered: The planar IC—revolution underestimated By Maury Wright, Editor in Chief -- EDN, 4/27/2006
Bob Noyce was at Philco Research in 1955. I had brief exposure to his small lab across the hall before he left. Philco at that time had the fastest pnp transistor SB100 that attracted a couple military computer contracts which I had responsibility. Philco disappeared from earth after Ford purchased it.
The rumor within Philco was management invested too much in jet etching production to put money behind Bob's planar approach. About 10 years ago, Moore responded to my inquiry that the planar approach was invented after Bob left Philco. Guess the internal Philco rumor was incorrect but sounded very plausible in the 1950s.
Sy Wong - 2006-3-5 07:49:00 PDT
