Posted by Tom Foremski - December 3, 2012
Ted Jenkins (circled) was personally recruited by co-founders Gordon Moore and Andy Grove, and once managed Paul Otellini, Intel's current CEO. This photo is from 1969 when Intel had just 106 staff, (104,000 today).
Robert "Ted" Jenkins was Intel's employee No. 22 and was hired away from Fairchild Semiconductor by co-founder Gordon Moore and Andy Grove in 1968. Now 69, Jenkins retired in 1999.
Over the course of his career, he worked directly for Grove, and future Intel CEO Paul Otellini reported to him during the early 1980s. Along the way he and the teams he led filed thousands of patents.
Paul Otellini worked for you in 1983 when you were GM of the Peripheral Components Division. What do you make of his recent retirement announcement?
I say congratulations to Paul. While at Intel, I used to say that one had to reinvent oneself every 5 years or so. It's harder to muster the same vigor to solve the same problem the 10th time than it is to solve it the first time. I'm pleased to having him join the rest of us Intel alumni. Maybe I'll be able to see him more often.
You patented a way to process a bipolar integrated circuit without using gold. How important was that IP to you being recruited by Intel from Fairchild in 1968?
It certainly helped. It was a process for an aluminum-silicon Schottky diode that could be used for what's known as a Baker Clamp to keep the bipolar transistor out of "saturation" so that it can operate at maximum speed. I knew it would be important for the digital circuits and we did decide to use it for the Intel bipolar process. This was the process for our first product the 3101, a 64-bit static RAM introduced in 1969.
When did you first meet Andy Grove?
I was introduced to him at my first interview. Back then he'd need to wear a headset because he always had a hearing problem. He asked a lot of technical, pointed questions and was very direct. It was a little daunting for a new college graduate.
[Later on] I remember having to do a monthly progress report when I worked under Andy. I used the word "corroborate" and he sent me a note, saying there's no such word. "You mean 'collaborate,' he wrote. I responded with my own note and told him, 'Corroborate' is a legitimate word."
He sent back one final note that said, "'Bastard' is a legitimate word, too."
Back in the 1960s, did you believe Moore's Law would hold true more than 40 years later?
The whole scenario blows my mind. Gordon proposed this in an Electronics magazine article in 1965, while he was director of R&D at Fairchild Semiconductor and I was between my BS and MS degrees at Caltech, a year before I joined Fairchild R&D.
As one who turned the crank on [Moore's] Law from the engine room, there were times when I thought the old man on the bridge might be crazy. The critical dimension was 10 microns when I joined the industry in 1966. It's now 22nm. That's an area compression of 250,000. That is an amazing journey.
By the time you retired in 1999, you and your program had filed around 2,000 patents. Are you still filing patents?
A little bit. I just had a patent issued that might help with global warming. Large volcanic eruptions cool the planet because aerosols from that reflect some of the sun's radiation. After Sept. 11, 2001, with all the U.S. aircraft grounded, we had a bump of about 1.5 degrees Celsius in the mid-U.S.
The patent covers the delivery of volcanic-like reflective materials from commercial aircraft. It could allow the use of cheaper, higher sulfur fuels, but the idea still requires testing and it is not without potential issues.
What technology do you think should have reached market by now?
Instant-on. I've always hoped that would happen. We're getting close to it with the MacBook Air and even closer with the Ultrabook. The Ultrabook almost qualifies as instant-on, but we're not quite there yet.
What do you think is your most memorable contribution during your career at Intel?
Probably starting Fab 3 in Livermore [Calif.]. This was the first fab where we wore bunny suits and I recruited most of my team from outside Intel. Many of them were early-career professionals from outside the industry who didn't know what they couldn't do.
In the era before PCs we were producing about 75 percent of the world's supply of DRAMs. In those days, they were used as cache memory in mainframe computers.
As your former employer looks for a new leader, what advice would you offer?
The next leg of the business looks like it depends even more heavily on human needs, interfaces and the management of stored and moving data. Strategically planning for this and executing as Intel has always done is the best path. The best of luck -- I'm still a stockholder.