Tuesday, May 20, 2008
I am doing some volunteer work at the Computer History Museum in Mountain View, CA and last Friday I had the fantastic experience of taking an oral history on video from Joe Costello.
The CHM is both the obvious - a museum on the history of computers – but also the non-obvious - the custodian of the history of the people behind the artifacts. The museum has been collecting oral histories from the people who created computer, semiconductor and software technology and the stories about the development of the companies which made the industry. The board of trustees hopes to capture both the stories of the technology but also the personal stories – the choices people made as they created such a vast impact on our society.
After the death of my close friend, Dean Richard Newton, we realized that we had lost one of the original pioneers of the EDA industry – and I was asked to be on the team of volunteers to capture history from the industry creators who are still alive. EDA (electronic design automation) is the industry behind the design of semiconductors – the incredibly complex software that make modern chip design possible.
Joe Costello was the CEO of Cadence as it grew from a tiny startup to the largest EDA company, and his stories are the stories of the formation of the EDA industry. His impact was larger than life and so it didn’t take much research for me to develop a timeline of questions to coax Joe through his memories of the creation and development of Cadence as we know it today.
Joe is a funny, animated guy so it took plenty of self control for me not to laugh or interact too much as he covered so many aspects of both his own career and EDA, only a fraction of which I can capture here.
- He studied mathematics at Harvey Mudd, then physics at Berkeley, but, after 2 masters, never finished his PhD because he enjoyed working in software at National Semi too much – it was just more interesting
- His first job at SDA systems (the startup that turned into Cadence) was as head of customer service – but it was an easy job because there were no customers – that was until they were on the ropes and he personally won a $20M partnership with Toshiba
- He was an accidental CEO not once but twice! The first time Jim Solomon (the founder of SDA) asked him to do it although the investors wanted to hire more experience; the second time Paul Huang of ECAD (which merged with SDA to form Cadence) asked him although the board wanted the more experienced CEO of ECAD. Today he just laughs at his luck and naivete at the time.
- He told the story of wanting to merge Cadence, Synopsys and Gateway – and proposing it at the craps table in Las Vegas. Had he succeeded he believes the industry would be quite different with one major player, but while Prabhu Goel (CEO of Gateway) wanted to – and did, Harvey Jones (CEO of Synopsys) didn’t and the rest is history. Cadence and Synopsys competed for the #1 spot then, and still do today.
- I hadn’t known that Cadence’s market leading analog products came out of Jim Solomon thinking of leaving. He felt he couldn’t contribute much more to Cadence’s digital strategy and so Joe asked him to take on a new field and gave him carte blanche. The result was Cadence had the best analog solution from then on.
- And there’s so much more… the Avanti theft of Cadence IP, the successful mergers and the failed merger with Valid, so many lessons. But what comes through loud and clear is how exciting the EDA industry was as it grew from a tiny fledgling to a mature market, how much Joe loved building Cadence over 13 years and how much passion he still has for the customers and solving their problems.
If you want to watch the colorful 3 hours of video we shot you’ll have to wait until the CHM gets it up on the web, hopefully soon, but if you haven’t been to the CHM I encourage you to go and see the videos of other pioneers of the computer industry.
And my next oral history will be Harvey Jones, hopefully in the next couple of months. Since I worked for Harvey on and off for 17 years I know too much, and yet not enough - I can’t wait.