Ed and Intel Investors - Intel Investments and Venture Capital
This is a rather lengthy, but revealing, look at Intel and its reasons for investing capital in other companies. Note the "real" cost of a 0.18 inch wafer fab, as discussed by Gordon Moore.
Paul
{========================================}
techweb.cmp.com
EDTN is a Member of
CMPnet
Posted: 8:45 p.m. EST, 2/13/98
Invest or die: Intel's life on the
edge
By Ron Wilson and Brian Fuller
SANTA CLARA, Calif. -- With about $600 million to
pump into venture companies this year, Intel Corp. has
joined the major leagues of venture-capital firms. But
the unique imperative that drives the microprocessor
giant to invest gives it influence disproportionate to
even this large sum. For Intel, venture investments are
not just a source of income; they are a vital tool in the
fight to survive.
Survival might seem an odd preoccupation for the
world's largest semiconductor company. But Intel, in a
way all its own, lives hanging in the balance. For every
new generation of CPUs, Intel must make huge
investments in process development, in buildings and
in fabs-an investment too huge to lose.
Gordon Moore, Intel chairman emeritus, gave scale to
the wager. "An R&D fab today costs $400 million just
for the building. Then you put about $1 billion of
equipment in it. That gets you a quarter-micron fab for
maybe 5,000 wafers per week, about the smallest
practical fab. For the next generation," Moore said,
"the minimum investment will be $2 billion, with maybe
$3 billion to $4 billion for any sort of volume
production. No other industry has such a short life on
such huge investments."
Much of this money will be spent before there is a
proven need for the microprocessors the fab will
produce. In essence, the entire $4 billion per fab is bet
on the proposition that the industry will absorb a huge
number of premium-priced CPUs that are only
somewhat faster than the currently available parts. If for
just one generation that didn't happen-if everyone
judged, say, that the Pentium II was fast enough, thank
you-the results would be unthinkable.
"My nightmare is to wake up some day and not need
any more computing power," Moore said.
With so much at stake, growth in demand cannot be
left to chance. Intel has learned to look down the
pipeline, from the development of a new process to
the architecture of the CPU built in it, the architecture
of the PC using the chip, the applications running on
the PC, the communications that link the PCs together
and even the distribution networks that will sell them.
At each of these stages, Intel acts aggressively to
prevent any obstruction from blocking the demand for
increased CPU speed. If the motherboard architecture
can't handle the bandwidth, Intel will invent new buses.
If core-logic vendors and motherboard suppliers don't
adopt the buses, Intel will build core logic or
motherboards. If applications don't need the
performance, or if users can't get enough network
bandwidth to use it, Intel will nurture applications and
seed broadband networking development.
Altogether, Intel has employed three complementary
strategies to pursue its ends. If it can, the company
seems to prefer to stimulate technology development
elsewhere in the industry. But when it feels the need for
more persuasive measures, Intel has developed or
purchased new technology-for example, the PCI and
AGP buses, new generations of cache SRAMs or,
more recently, the Rambus-Direct DRAM and the i740
3-D graphics chip. This new technology is frequently
made available to others to manufacture. If even that
doesn't get things moving in the direction Intel needs,
the company has shown itself willing to intervene
directly: moving into the core-logic or motherboard
business, for example, or productizing the i740.
Most of Intel's activity, however, is not so
heavy-handed as going into the graphics-chip
business. For the most part, the company has been
able to pursue its objectives by nurturing other-usually
small-companies with money, technical information
and the might of the Intel name. That strategy is the
responsibility of Leslie Vadasz, Intel senior vice
president.
"There are probably 2,500 companies that contribute
to this market," Vadasz said. "Our job is to make sure
those companies have the technical and financial
resources to pursue those markets."
This mission has put Vadasz in charge of a far-flung
investment empire. Reaching into every stage of the
computer industry's pipeline, from semiconductor
process equipment to PC retailers, Vadasz uses
Intel's money, prestige and technology to nurture-and
to some extent direct-the fate of venture companies.
The size of Intel's purse, and the size of its reputation,
have made it a presence in the VC community. "We've
come across them a number of times," said Sam Lee,
general partner with information Technology Ventures
(Menlo Park, Calif.). "They told us they invested $300
million last year and intend to
double that this year." Lee described the investments
as small and strategic-"$250,000 here, $300,000
there."
In many deals, Intel wields a huge stick that even
venture capitalists can't. For instance, it can, just
because it's Intel, demand more stock than would
otherwise be given for a certain level of investment.
The company brings to the table not only cash but
cachet.
"They can make or break you in many ways," Lee
said.
Intel has to be aggressive in its investments in a way
that Microsoft Corp. doesn't necessarily, said Mark
Dubovoy, another general partner at ITV. Intel has to
spread its technology and still make $600 chips.
"Maybe they take small positions in these companies
and say, 'you need more graphics, more speed.' "
Still, Intel is no Genghis Khan. "They don't dominate.
They're supportive and they don't meddle," Dubovoy
said.
More often, other players see Intel as an investor and
advocate. Vadasz agrees that Intel makes clear its
reasons for investing in a new company, and
encourages that company to act in ways that will, in the
long run, increase the need for CPU cycles.
Vadasz, who was employee number three at Intel, is
no stranger to the advocacy role. When the
corporation identified the lack of broadband
connectivity to homes as a roadblock to continued
PC-market growth, Vadasz took the helm of Intel's
Telecom Policy effort, one of the more intensive-and
frustrating-lobbying efforts undertaken by the U.S.
semiconductor industry. Now, he manages a portfolio
that includes over a half billion dollars invested in
better than 100 companies.
"We have two strategies," Vadasz explained: "one to
create new market segments, and the other to enable
us to do a better job of shipping our products. Under
the first strategy, among the many subsets of the task
of moving the PC into new areas, we have been very
interested in expanding content and expanding
bandwidth."
For content, Intel currently holds investments in about
eight or 10 "new media" start-ups. In general these
companies are exploring new ways of using sound
and graphics as a human interface to business and
industrial tasks-data mining, for example, or medical
diagnostics, or even network navigation.
Such companies could potentially address a crucial
problem Intel faces in the market. To sell more CPU
cycles, the company must see growth in the use of 3-D
graphics. But to date, important applications for 3-D
have emerged only in the relatively small market of
high-end games and the even smaller one of
computer-aided design. Intel needs 3-D to become a
mainstream human-interface tool.
The company is also investing in businesses that
could increase the graphics content, traffic level and
penetration of the Internet. Many of those investments
are in electronic commerce. Similarly, Intel is investing
in ideas for the notoriously slippery, but unquestionably
growing, small office/home office (Soho) market.
For any of these markets to spread beyond the mature
commercial desktop, Intel realizes, networking
bandwidth to the home must improve. It is unrealistic to
expect virtual-reality shopping, networked Soho
workstations and the like on a twisted pair that rarely
achieves better than 24 kbits/second. "There are 45
million homes in the United States with PCs," Vadasz
observed, "but less than 10,000 with broadband
connections."
Intel initially counted on its lobbying efforts to break the
logjam, but it seems increasingly pessimistic about the
prospects for a legislated improvement in home
connectivity. "Naively, I thought that the Telecom
[Reform] Act would have a major impact on broadband
service to homes," Vadasz said. "But . . . the act had
no real impact on broadband; it was just encouraging
competition on plain old telephone service. We've
been talking to anybody who would listen, increasingly
in Congress. But I doubt that anything will happen in
the way of new legislation in 1998. The policy of the
government is to deregulate, not to deploy."
As Plan B, Intel has begun spreading investments
among companies that could offer alternatives to telco
service. It has poured money into novel wire-line
techniques, cable data technology-Hybrid Networks,
for instance-and satellite service, including a position
in the European Astra broadband data-broadcast
system.
The second leg of Intel's investment strategy is
perhaps the more obvious: sinking money into
technology to design and build more CPUs. Here the
company has cast its net wide, funding work in
development tools, process equipment and other
areas.
One example is the emerging area of
formal-verification tools. Intel identified formal
verification as a need several years ago and has met
its own needs through internal development and
external investment.
Another focus is lithography. "One example of this is
the extreme-ultraviolet stepper program," Vadasz
said. "At one point, EUV was a Department of
Defense program. We became interested and later
began to believe that EUV could be practical. When
the program went on the budget chopping block, we
stepped up with a relatively large investment. I think we
saved a program that could delay the need for X-ray
lithography, and we could make a good profit on it as
an investment."
That duality governs all of the investments Vadasz
makes. They have a strategic goal, but are also
expected-as a secondary consideration-to be
profitable. "No matter how great a technology is, if the
company we invested in doesn't succeed, they haven't
done a damn thing for our strategies," Vadasz said.
That duality leads to a complex, but carefully managed
relationship between Intel and its investments. While
Vadasz and his team manage the business
relationship, the investee also will have a strategic
champion from somewhere within the engineering
ranks at Intel. The actual relationship can be relatively
arm's length.
"Sometimes we just periodically call to see how things
are going," Vadasz said. "Other times, we ask to have
an Intel manager as an observer on their board. If
there's a problem between us, we will let our feelings
be known, but we don't have a vote. We don't want to
get involved in running other people's companies.
That's one reason we tend to work closely with venture
capitalists in funding these companies."
Yet Intel's influence can be disproportionate to its
investment. It's not just that Intel is the archetypal
800-pound gorilla. The company has another gem to
offer investees, one of inestimable value that is
available nowhere else.
That precious resource is information. Often-though
not always-an Intel investment will be accompanied by
a technology-exchange agreement. A new-media
company, for example, might get trade-secret
information about how MMX-2 will work. Or a graphics
partner like Lockheed-Martin's 3-D operation might
get early specs for the AGP bus. Such information can
give even a small company an insuperable advantage
in breaking into a competitive market.
"It can be more important than the money," said one
new-venture manager who wished not to be named.
"To know what Intel is going to do in the next
generation, maybe a year before your competitors
know, is an incredible advantage."
Such a valuable gift raises very complex issues for
Intel. Under the law, the funds must not be used in a
way that would restrict competition unfairly. "We do
have development partnerships with some companies,
whether or not we invest in them," Vadasz said. "But
we have to be very careful when giving information to a
company we invest in. There are no hard rules, so our
analysis of what we can do has to be very situational."
The issues become even more complex when the
company is in another country, with another culture.
"The strategy definitely has to be adjusted for different
regions," Vadasz said. "We have to take a look at
what's happening in the local markets. Sometimes
what we need to invest in is not the technology, but the
channel."
In the People's Republic of China, for instance, Intel
has invested not only in technology-related companies,
but also in distribution-channel companies. In that
market, simply delivering PCs to their users was a
challenge that needed attention.
There are other aspects of investment in a country like
China that Intel has had to watch closely. Some of the
infrastructure there is owned, directly or indirectly, by
less-than-ideal business partners. One such owner
that raises particularly touchy issues for a U.S.
technology company is the People's Army. "We try to
be very careful how we handle the People's Army
issue in China," Vadasz said. "We want to avoid
situations where we feel we might have to
compromise ourselves."
Through this labyrinth of complex issues Intel continues
to steer its growing investment program. A company
that even at the pinnacle of success must look
nervously over its shoulder, Intel sees opportunity not
only in making money on new ventures, but in helping
them to solve its own long-term problem: the perpetual
need to increase demand for CPU cycles. |
