Computer pioneer Gene Amdahl, who died last night at age 92, is being remembered today for his pioneering contributions to computer architecture, as a key developer of the IBM System/360 mainframe computer and the Amdahl 470. He’s also known for Amdahl’s Law, a theory used in parallel computing. For me, he was a big part of my introduction to Silicon Valley, and he spent hours back in 1982 trying to help me understand why all these engineers were starting companies, why it was OK to fail, and how to get up again and start over when you did.
I had come to Silicon Valley from New York City on one of my first West Coast reporting trips. My assignment—talk to some of these EEs who had taken the entrepreneurial path and try to find out what made them tick. I visited 14, including Steve Wozniak, Adam Osborne, Gary Kildall—and Gene Amdahl. Amdahl had recently left his first startup, Amdahl Corp., having lost control to investors—a familiar Silicon Valley story.
Here’s what he told me at the time, as I reported in the July 1982 issue of IEEE Spectrum:
“Our original finance plan for Amdahl Corp. included a public offering in late 1973, but that was at the deepest point in the recession, so we had to raise money from private sources. In 1974, through a large investment, local venture capitalists and a Japanese company assumed the controlling interest. The U.S. investor perceived the happiness of the Japanese investors as most important, so more and more of Amdahl’s functions were taken over by the Japanese company, until Amdahl could no longer do them itself.
“I was so frustrated by the directions the company was taking that the muscles in the small of my back spasmed. I resigned my position at Amdahl, and in September 1979 I went into the hospital.
“Instead of taking pain pills, I put my mind on problems that had to be resolved to make great advances in the computer field, and found that seven months of concentrating paid a lot of dividends—so much so that I could not resist trying to put them into practice.”
So he started Trilogy Systems Corp. in 1980 to design a chip intended to allow cheaper mainframe computers to be built. The effort failed, but it wasn’t his last startup, or his last failure. He eventually had a limited success with a company that started as Commercial Data Servers but later changed its name to Xbridge Systems. It creates data mining software to scan for personal information in large databases and protect it.
Why start companies, instead of just design computers?
“Engineering,” he told me, “is satisfying people’s needs, and people’s needs aren’t all technical. Engineering is really a social undertaking.”
Tekla S. Perry is a senior editor at IEEE Spectrum. Based in Palo Alto, Calif., she's been covering the people, companies, and technology that make Silicon Valley a special place for more than 40 years. An IEEE member, she holds a bachelor's degree in journalism from Michigan State University.