After
his bachelors in Electrical Engineering from the
Delhi College of Engineering, in 1971, Dham wanted
to go abroad and study microelectronics. However,
his parents wanted him to be with them in Delhi.
" At that time, career and such things did
not enter my mind. There was simply no protest,
I accepted it," recalls Dham. Yet, he was lucky
to be 10 minutes away from a forward-looking entrepreneur,
Gurpreet Singh of Continental Devices, who wanted
to run a world-class semiconductor company in the
outskirts of Delhi. "Though Sardarji, as we
used to call him, was not a techie - he was an economist
from the London School of Economics - he was very
aware that in semiconductors lay the future. He
was collecting bright people from places like Berkeley
and Stanford. He had met Gordon Moore. Robert Noyce
had even come and stayed in his house in Maharani
Bagh, in the late '60s, since Noyce wanted Intel
to start chip manufacturing in India," says
Dham.
Dham
wanted to know what went on inside the devices.
And so, after convincing his parents, he went to
Cincinnati in 1975 to do an MS EE in Solid State
Sciences. Cincinnati, at that time, was a very good
school in microelectronics with even a fab on campus
and was widely supported by the semiconductor industry.
After
MS Dham went to Dayton and joined MCR. "It
was cold and lonely but I got my green card and
work experience. I got some patents from the work
I did there. I was presenting it in a IEEE conference
in Monterrey, California. The Intel people were
also there presenting their work and they said they
wanted me to join them. While in Continental Devices
I had read about Gordon Moore, Robert Noyce and
all that in the technology magazines but they were
asking questions like how PhDs would run a business.
It was fascinating. Of course, I joined as a lowly
engineer. I worked on EPROMS (I was a co-inventor
of flash memory) for seven years," recalls
Dham.
However,
Dham wanted more action and started looking around.
At that time, the 386 chip had been designed and
had gone for production. Dham wanted to get into
microprocessors, he applied for a job in that division
but he was rejected since the project was on course.
That would not deter a determined Dham. He went
nosing around and found that there were problems
in production. The fab thought may be it was at
fault and was cleaning up its shop, the designers
were at tethers end after several redesigns and
Dham thought he could lick the problem. He went
to the programme manager and told him that he would
act as his consultant and need not be given a formal
position. When Dham straightened out the problem,
Intel's fortunes shot up and the boss was happy.
So he made him in charge of 386 and went on to 486
himself. "But 386 was now mature and there
was not much excitement." It so turned out
that he was then shifted to 486 since his former
boss had quit. However 486 was in deep trouble,
the fab was ready, the chip had been announced to
go one up on competitors but there were numerous
problems at all levels. "I worked so hard I
thought I died, but finally I finished the project
in November 1989. I took a month off to India to
unwind and came back in January 1990 and was made
incharge of 586 or Pentium," says Dham.
Pentium
was a challenge in many ways; 486 was more integration
than innovation. Paranoia was absolutely at the
top. "It was not easy at that time. The first
six months was foundation laying. We also started
bringing multi-scalar architecture. I picked Avtar
Saini to execute the design. He was a go-getter
and executed very well. The whole team did a great
job. The biggest challenges came in business. The
big customers like IBM and Compaq were very happy
selling 486. But we wanted to stay one generation
ahead of competition. But, our customers were not
ready. Luckily CD-ROM prices crashed and became
affordable for home PCs. We had a bus called PCI
and we could put graphics, audio and video as well
as games on a home computer using Pentium, and then
position ourselves ahead of 486. We picked ourselves
a new horse called Packard Bell, which nobody in
the corporate market had heard of. They started
selling to CompUSA, Circuit City, Best Buy, Good
Guy and all the retail stores. I used to go with
my team to demos with 486 and Pentium to show how
Pentium was better than 486. They were getting it
only from Packard Bell and they asked Compaq and
Dell: "Where is your Pentium machine?"
Pentium became a huge success for Intel and Dham
left Intel in 1995, riding on its success.
"The
best thing that happened to me was joining Intel
and the best thing that happened to me was leaving
Intel," says Dham in one of his crisp sound
bytes that make him so popular with journalists.
He
joined Nexgen, which was a startup that was acquired
by AMD later. After helping AMD seriously challenge
Intel with its vastly popular K6, Dham left AMD
and joined Silicon Spice, a startup, as chairman,
president and CEO though others had founded it.
"It has been the best part of my life, building
teams, products, raising money, talking to customers
and finally selling it to Broadcom, a company which
might become tomorrow's Cisco," he says. Silicon
Spice has been acquired by Broadcom for $1.2 billion
and everybody, including some office staff, have
become millionaires.
Photographs
and certificates from Andy Grove and Craig Barret
about 386, 486 and Pentium adorn Dham's office walls
as well as one from Bill Clinton for being the presidential
advisor on minorities. Noticeably his latest chip,
Calisto - its very first copy that passed all tests
- lies at the feet of a small Ganapati statue on
his table.
Dham's
favourite hobby is carpentry and his favourite TV
show is Home Improvement. 'Tool Man' Tim Taylor's
Do It Yourself does not quite work. This hi-tech
craftsman's chips sure do.
ARTICLE
SOURCE : DELHI
COLLEGE OF ENGINEERING, UNIVERSITY
OF CINCINATTI
