...The Gartner Group's Neil MacDonald says Lucent and Nortel Networks are at least a year ahead of Cisco in key aspects of QOS technology...
forbes.com
Can Cisco do in telephony what it has already
done in computers? Judy Estrin has one helluva
challenge.
Wired
By Mary Beth Grover
THE GREAT TELECOMPUTING convergence is in
full swing: Computer networks converge
headlong on telephone networks, creating and
destroying businesses along the way. Cisco
Systems, the leading power in the computer
networking business, must extend its reach into
telephone networks or lose momentum. The
responsibility rests heavily upon Judith Estrin,
Cisco's 44-year-old chief technology officer.
Were it not for her indifference to publicity,
Estrin would probably be one of America's
best-known female entrepreneurs. A Stanford
graduate with master's degrees in both electrical
engineering and computer science, she was just
26 when she cofounded the first of three startups
and hit a home run with each.
Two years ago FORBES featured on its cover 20
distinguished women from Silicon Valley
—Estrin was the only one not shown. "I'd rather
be with my customers," explains Estrin, who is
small of stature and favors plain skirts and
blouses. She hardly exudes charisma.
Despite her low profile, Estrin stands out. She
and her husband, William Carrico, started Bridge
Communications in 1981 and merged with
3Com for stock worth $235 million. Her
Network Computing Devices, started in 1988,
went public four years later. Last summer she
sold Precept Software, a multimedia company
founded in 1995, to Cisco for $84 million in
stock and went along as part of the deal.
As chief technology officer for Cisco, Estrin
presides over a $1 billion R&D budget and an
ambitious acquisitions strategy. If Estrin is a
phenomenon, so is Cisco. Only 14 years old, this
San Jose, Calif.-based firm enjoys a stock
market value of $123 billion. It makes routers
and switches that enable computers to talk to
one another, and dominates this business: Eighty
percent of the world's Internet traffic goes
through Cisco equipment. A couple of years ago
we predicted that Cisco would plateau as routers
lost ground to cheaper switches (July 29, 1996);
we were wrong. Since then its revenues have
more than doubled, to $8.5 billion.
Now Cisco is moving from the Internet into the
related business of routing voice and data traffic
over the world's telephone network. There it
confronts heavyweight competitors such as
Lucent Technologies and Nortel Networks.
There is a fundamental difference between
computer networks and telephone networks. In
computing, innovation is the most important
thing. In telephony, reliability is. Estrin has the
formidable task of proving that Cisco can deliver
reliability as well as Lucent or Nortel can.
Think about the computer you use at your office.
There are so many hardware and software
innovations that you can't keep up with them all.
But all that innovation begets instability. If you
suffer a crash once a week, you are doing no
worse than most computer users.
Unimaginative the phone company may be, but it
does know something about bulletproofing its
equipment. Can you remember the last time you
picked up the phone and didn't get a dial tone?
And yet telephony and computing are close
electronic kin. They both work almost
exclusively in the digital 0s and 1s of computer
language; inside a telephone central office are
computers running a few million lines of software.
That giant hunk of software doesn't do much
except route calls—and allow the local phone
monopoly to provide some overpriced features
like caller I.D.—but it does so with extreme
reliability.
This is what Estrin has in mind when she
describes Cisco's new thrust. "We can't just
think about routers and switches anymore," she
says. "It's software, software, software."
Software advances, she explains, will enable
Cisco to build better firewalls, which will keep
intruders and viruses from crippling computer
networks, and to make its routers more
"intelligent."
Intelligent in what way? "I really can't talk about
those things," she answers, adopting the caginess
that comes naturally to an engineer from a
startup.
For an explanation we turned to one of Cisco's
customers. Says US West's Robert Zell:
"[Cisco] has the equipment to compete in the
voice market. But they need quality of service."
Quality of service—QOS, in telco jargon—is
getting to be an often-heard theme. It refers not
just to reliability but to software that enables a
switch or router to set priorities. The equipment
must distinguish different kinds of data and give
priority to, say, voice or bank ATM transmissions
over less important e-mails. You wouldn't want a
conversation held up in the network just because
someone in accounting is surfing the Web.
Quality of service features can help eliminate
those irritating long pauses you get when phoning
someone over the Internet, and they can turn
herky-jerky video images into sharp pictures.
With QOS, telcos can offer customers various
services at different rates—discounts, say, for
service with pauses tolerated or without video,
which causes uneven bursts of bandwidth
demand.
Cisco has spent some $7.3 billion in the past five
years acquiring 30 companies, trying to fill in
some huge gaps in technology. Estrin has to
make the pieces work together—and she does
not have a lot of time. Nortel and Lucent are
rushing to provide QOS, and so are a handful of
scrappy upstarts in telecomputing gear, such as
California-based Ukiah Software and Packeteer.
The newcomers, moreover, are pushing QOS
software and other intelligent features not
necessarily on routers and switches, where Cisco
holds sway, but on servers, where it does not.
The Gartner Group's Neil MacDonald says
Lucent and Nortel Networks are at least a year
ahead of Cisco in key aspects of QOS
technology.
Estrin, however, knows from direct experience
how quickly a lead can evaporate in this
business—and she doesn't lack for
self-confidence. She and Carrico launched
Bridge Communications three years before
Sandy Lerner and Leonard Bosack, another
husband-and-wife team, began Cisco.
"Could we have been Cisco? Yes," says the
cocky Estrin. "We probably would've bought
them." But Estrin left Bridge after quarreling with
the boss of its new owner, 3Com. He apparently
didn't much like being told by a 32-year-old
woman that networking was the future—not
computers based on IBM's OS/2 system.
Two years ago we underestimated Cisco. This
time around, we wouldn't bet against Estrin—or
Cisco. |
