December 11, 2019

Accidental Empires, Ch. 3 - Why They Don't Call It Computer Valley


Cringely stresses that constant development characterises the PC industry whereas mainframe technology lasts 20 years (cf PCs change every 18 months).

Was there a transition from mainframe to PCs? Cringely says no. 'Personal computers have almost nothing to do with big computers. They never have.' The mainframe industry is more than building and selling a computer . After sales is equally lucrative - servicing, financing, programming. PCs sell to millions, mainframes to a few.

The developers from the two industries are different. The PC industry grew out of the semiconductor industry. The PC is a "big chip", not a small mainframe. Silicon Valley is not called Computer Valley.

William Shockley

Invents transistor, Bell Labs late 1940s, forms Shockley Semiconductor.

Bob Noyce

Noyce plus seven others split to form Fairchild Semiconductor, 1957. Template for all chip development. Internal dispute leads to a group of engineers splitting with venture capital for start up.

Company ethos matches its founders. High tech "regular guys" , no social hierarchy, informal. Noyce fostered creativity and openess, "crazy ideas".

Noyce left Fairchild to form Intel (wanted stock options for all staff, another dispute).

Semiconductor business has high development costs and comparatively low manufacturing costs. "A volume business". Noyce pioneered high volume low cost, continually cutting unit price to build volume. To avoid increasing manufacturing costs Noyce developed integrated circuits IC, building more and more components into a single chip, cutting down on assembly labour.

Moore's Law

Gordon Moore, co-founder of Intel. The number of transistors that can be built into a chip doubles every 18 months.

IC deveopment led to more complex chips and hence ability to sell at higher prices.

Ted Hoff

At Intel, 1971, develops microprocessor, programmable chip. When memory and support chips were added produces a computer.

Computer companies (IBM) didn't want microprocessors for mainframes, too cheap which would undermine their price structuring. They didn't see their potential. The next generation of chip engineers did. They built computers and followed the template. Volume, increasing functionality, continual development and decreasing price.

Cringely has an interesting theory. He says that it takes 30 years for society to absorb a new technology into everyday life. So if the early 70s mark the birth of the PC then we are about to hit the point when we are going to find a real use for PCs.

"..technologies found their greatest success being used in ways other than were originally expected."

Early development of PCs was by enthusiasts, amateurs, fed by a chip industry looking to find ways of using their new products and by mainframe programmers. Combination of counter culture and disaffected corporate workers. With the notable exception of Bill Gates, Cringely claims that innovation and an enthusiasm for the machines, and consequently the development of software to run them, was the spur, not money.

Next page » Intel

Previous page « Types of computer




















Up to top of page