Intel changes its research agenda

Intel Corporation’s leadership in the microprocessor industry is
undisputed. The popular notion is that Intel’s research efforts are solely
directed towards packing in more transistors onto the silicon real-estate,
devising new architectures that speed up processing, innovating on semiconductor
packaging and manufacturing and now producing chips and chipsets that enable
wireless computing.

Having more than a dozen processors under development with various code-names
and at different stages in the roadmap, it only seems plausible to conclude that
Intel’s game rests at retiring old generations of processors and introducing
new ones with stronger specs, backed by high horsepower marketing. More and
better and yet cheaper (the world calls it Moore’s Law) drives Intel.

But as poet Robert Browning said, ‘the best is yet to be, the last of life
for which the first was made’, Intel has embarked on a new course as its
research suggests. It is focusing its research efforts on a whole set of new
areas that may seem very unconnected and disjointed from what the company is
now.

Intel showcased some of its leading edge research activity to a select
international media at Santa Clara, CA last week. The scope was breathtaking and
the areas diverse : silicon photonics, compu-tational nanovision, context-aware
computing, personal server, sensor networks, instrumentation systems, RFID,
optical systems, myriad sets of new devices and even ethnographic research on
how people compute in specific countries and emerging scientific disciplines
like precision biology.

Not that Intel is planning to enter all of these areas in the future but the
diversity is needed to spot the right opportunity it can exploit to innovate and
gain leadership.

In an exclusive interview with CyberMedia News, Pat Gelsinger, Sr VP and CTO,
Intel, said, "Some of the areas that you see here are specula-tive and we
may not go into these areas but we spread ourselves wide and deep that no
opportunity escapes us and we get to create new markets."

As the CTO, Gelsinger is concerned about the US losing its edge in pure
research activity and trading it off for commercial success. He said, " I
have spoken about this in public, as a country we have been making bad policy
decisions". He is encouraging a meaningful shift in this case- for example–he
managed to get the National Science Foundation (NSF) fund doubled, Gelsinger
added.

Not that Intel wants to center its research in the US. As a worldwide
supplier of technology, which is a product of research, Gelsinger opted, "
If there is a better research agenda somewhere else- say China or India, we will
pursue that." That explains Intel’s ethnographic research on countries
like India, Malaysia and Brazil amongst others.

Said Gelsinger, "We have to do culturally relevant research to be able
to sell products in that country tomorrow." Most of the work that Intel
showcased was not about semiconductors. Does that mean that Intel is shifting
gear? Explained Gelsinger, " Much of the work required is non-silicon.
However, at the end of the day, we will embody all of this in silicon."
This is because Intel has sort of mastered the semiconductor discipline and has
huge investments in fabricating them.

Intel’s fabs have gone bigger and bigger but the net cost of manufacturing
chips have continually decreased. This becomes a entry barrier to others who can
do the research but can’t afford to build products that it can mass produce
and sell- whereas Intel has the best of both the worlds. Now, that’s the Intel
way.

Iishwar Daas

Santa Clara (CyberMedia News)