The transistor turns 60
Bell Labs scientists cracked a technical problem 60 years ago, and online-dating services are the result.
Correction, 10:45 a.m. PST: This blog initially misstated Fred Terman's title at Stanford University. He was provost.
Sixty years ago, on December 16, scientists at Bell Labs--William Shockley, John Bardeen, and Walter Brattain--built the world's first transistor and nothing has been the same since. We'll be covering the anniversary in subsequent articles, but here's a smattering of some of the implications, in somewhat chronological order, of the event:
1. The dawn of electronics. Vacuum tubes consumed lots of power and were fragile. ENIAC, one of the world's first computers, weighed 28 tons, consumed 170,000 watts of power and required several operators. It conducted 5,000 operations a second. Since the 1930s, Bell Labs had been looking to replace tubes with an electronic switch.
2. The birth of the insane boss. Technically speaking, Bardeen and Brattain invented the first transistor, a point-contact transistor. Shockley, who had been researching the problem for years, came up with the junction transistor, which became the basis of commercial transistors. Brilliant, imperious, and arrogant, Shockley ended up getting most of the credit. (Another Bell Labs employee, John Pierce, came up with the name.)
3. Silicon Valley. There's a good reason that the ISSCC conference, one of the premier confabs for chip designers, that takes place in San Francisco is organized by the University of Pennsylvania. The computer industry (think Sperry) used to be located back East. Fred Terman, then provost of Stanford University, started to recruit people like Shockley, who then recruited people like Robert Noyce, Gordon Moore, and Eugene Kleiner, to Santa Clara County. The $2 million dollar ranch house followed.
4. Predictable progress. One of the remarkable aspects of electronics is that progress is made at a steady, predictable rate. Things get cheaper, faster, and smaller over time. You can't say the same thing about the chemical industry, or pharmaceuticals. If the auto industry followed Moore's Law for even a decade or two, a Rolls Royce would cost less than a dollar and be far faster than the models on the road. (Granted, it would also be less than a centimeter long, but you have to accept some trade-offs.)
Making the future foreseeable led to:
5. Venture capital and a booming market for start-ups. No one wants to risk putting $2 billion into a fab, or even $20 million into a new search company, unless he can anticipate a payout. The predictability of electronics reduced the risks for investors, which in turn freed them to fund "crazy" ventures like Fairchild Semiconductor or Intel. Without predictability, most of the people I know wouldn't have jobs. Which in turn has led to:
6. An interconnected world. Cell phones, PCs, social networks, the Internet. If transistors didn't get continually faster and cheaper, the cell phone industry wouldn't be shipping a billion handsets a year. Every few years, some knuckleheads try to declare that Moore's Law, which charts electronics progress, is dead or irrelevant. And they usually research those predictions by going to search engines, which wouldn't exist if servers weren't plummeting in price.
Moore's Law, as it applies to planar silicon, will likely peter out around 2020, but chances are some new materials or structures will be invented that will let progress continue.
So take time out this month to thank the millions of tiny microcircuits in your phone.