|
Previous | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | Next
Again they turned to Arthur Rock. Noyce and Moore each kicked
in $250,000 of their own money and Rock raised an additional $2.5
million; Grinnell College, where Noyce was an active alumnus, invested $300,000.
Intel was incorporated on July 18, 1968, as NM Electronics (for
"Noyce" and "Moore"). Rock was Chairman of the Board; Noyce was
president and CEO, and Moore was executive vice president. They set
up shop in Mountain View, California, just down the road from both
Fairchild Semiconductor and Stanford University. Recruiting about a
dozen employees from Fairchild, including Andrew Grove, they set out to fill a niche but ultimately created a new industry. "The semiconductor memory business did not exist," said Noyce. "That's key to the survival of a young company. You try to go into a business that is either underpopulated or not populated."
Even if semiconductors could be made viable for memory storage,
others in the industry predicted that they would cost about ten times as much as magnetic cores. As a result, few firms saw any commercial possibilities in developing them. Intel (the company was renamed soon after the founding) intended to change this state of affairs by continually cutting production costs while cramming ever more transistors on a single chip. Within a few years, if Moore's law held true, memory chips would become cheaper and more desirable than magnetic cores. That would constitute a scientific achievement, but Intel was a business and had to establish itself in a nascent market. "We figured we had about five years to get established before the big semiconductor companies would follow in this market and become direct competitors," Moore recalled. One of the first decisions that the management team had to make was what they called the "degree of difficulty." If they produced a very simple product, others could easily copy it. If they tried for an
overly complex one, the company's resources might give out before the research was complete. In the end they chose a middle path, and predicted that within five years their new firm could have annual revenues of $25 million.
Intel had a long way to go. In its first year of business it reported negligible revenues of just $2,672. After a few false starts, Intel's scientists began to focus on producing the silicon gate metal,oxide semiconductor (MOS) in 1969. "We chose a technology that was tractable enough so that, by concentrating all our energies, we could get past unforeseen difficulties," said Moore. In 1970, Intel brought out its first successful product, the 1103 chip, which contained lK, or a thousand bytes, of dynamic random access memory (DRAM).
DRAM, despite its name, was largely passive. Information could
only be stored on it. Intel's next step was to make chips that were more than simple receptacles. The company achieved this goal, in part, with a second memory product that was developed simultaneously. One researcher, Dov Frohman, devised a chip that, like the DRAM, could store data permanently. It could also be erased, and therefore could be reprogrammed. The EPROM (erasable programmable read,only memory) chip was a quick, cheap, and easy way to store not just data but the programs that could give instructions to DRAM chips. Frohman recalled, with enormous understatement "We weren't geniuses. Invention is just
a process of dreaming a lot and then asking, 'Why not?' " The EPROMs helped boost the market for Intel DRAMs, and the company's sales rose to $9.43 million in 1971. The same year, Intel further bolstered its financial standing through an initial public offering that brought in $6.8 million.
Although the first products were major accomplishments, Intel's managers realized the company was far from realizing its goal of $25 million in annual revenues. "A lot of things are technologically possible, but only economically feasible products will become a reality," said Noyce.
From the start, the manufacture of silicon chips was complicated.
In the early seventies the factory would reduce a design through photography, and then imprint it on a tiny sliver of silicon. The process was repeated time and again to pack thousands of transistors on a single chip. Production of the chips was enormously expensive, and technological breakthroughs would have languished if Intel didn't devise ways, at every stage, to produce chips at affordable rates. Andrew Grove stepped up to do just that. "Noyce and Moore were the inspiration. Grove created the organization that executed," recalled Dun Hutcheson, an executive at the computer firm VLSI Research.
Grove, who had a mind for industrial organization, was put in
charge of production and helped to direct the company's initial experimentation with assembly lines. "The [fabrication] area looked like Willy Wonka's factory, with hoses and wires and contraptions chugging along," Grove recalled. "It was state-of-the-art manufacturing at the time, but by today's standards it was unbelievably crude." It worked well enough to make the chips en masse, substantially reducing unit costs. And since Intel had so little competition, it was able to charge a premium price. The company's profit margins soared.
Previous | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | Next
|