Doug,Re: Q: What will the network of the future look like?
Q: What will the network of the future look like?
A: There will be copper and there will be fiber, there will be
fixed radio and mobile and satellite, and each one will fill its
own niche. On top of these there will be a variety of protocols.
No one size fits all. There will be a lot of complexity, and your
machines will attach to whatever is out there. You will see the
emergence of things like bandwidth-on-demand, where you can
share a pipe and get the bandwidth you need. But Internet
Protocol [ the software lingua franca of the Net ] isn't going to
solve every problem. Not all of this stuff is going to be handled
by a single goddamn Internet Protocol network.
quantumbridge.com
Business Week Online
November 22, 1999 Issue
No, Virginia, the Net Is Not Going to Make
Everything Simple
Arno Penzias sees a multiplicity of e-devices in which 'no
one size fits all'
After 36 years at Bell Laboratories, including three years as
director of the legendary research center, Nobel laureate Arno A.
Penzias has retired in California. There, he serves as an adviser
and investor with blue-blood venture capitalist firm New
Enterprise Associates, offering his insights into
telecommunications and other technologies. Dressed in a red
sweatsuit in his San Francisco home, Penzias shared some of
those insights with BUSINESS WEEK's Andy Reinhardt, while
cargo ships plied the bay outside his picture windows.
Q: Why did you leave Bell Labs?
A: I had changed everybody's job except mine. So I decided to
change, and right at that point, I began to see all these little
companies doing interesting things out here. The work we did at
Bell Labs had set the stage for all this wonderful stuff. So I came
out here, initially with Lucent (LU), and started working with
small companies. Now I'm on my own and working with New
Enterprise Associates.
Q: What's your vision of how the communications system is
being transformed today?
A: There is going to be intelligence everywhere in the network,
but there will be considerably more control at the edge than there
is now.
Q: What is the difference between intelligence and control?
A: Intelligence is what allows a function to be carried out.
Control is where the choice is made to use the function. There
are big religious arguments about this. But the trend is
undeniable. It's like the Internet--where users have
control--compared to the old phone system, which was
completely centralized. This is a growing theme throughout our
whole society, and not just in the communications sector.
Q: Tell me more about decentralization.
A: Back in the industrial age, the image of progress was
Pittsburgh, with its huge stone chimneys belching smoke. Now
it's small offices and home offices. People can carry their
laptops everywhere, and those are their offices. The poster child
of American prosperity today is somebody with a laptop getting
on an airplane. For a telecom company like Lucent or Bell South
(BLS), what that means is that their job is moving from the
central telephone office to the customer premises. They're in the
best position to take care of networking stuff in the home anyway.
After all, they make house calls; they're the people who come
and make your phone work. Maintaining things like home
networks will become the locus of tomorrow's communications
companies. If the local telephone companies ever become
well-managed, they could be really dangerous [laughs].
Q: What will the network of the future look like?
A: There will be copper and there will be fiber, there will be
fixed radio and mobile and satellite, and each one will fill its
own niche. On top of these there will be a variety of protocols.
No one size fits all. There will be a lot of complexity, and your
machines will attach to whatever is out there. You will see the
emergence of things like bandwidth-on-demand, where you can
share a pipe and get the bandwidth you need. But Internet
Protocol [the software lingua franca of the Net] isn't going to
solve every problem. Not all of this stuff is going to be handled
by a single goddamn Internet Protocol network.
Q: Who said it would be?
A: Well, John Chambers [CEO of Cisco Systems (CSCO)] says
the telephone is a dinosaur. He's a great man, but his mentality is
that you should get rid of your phone and use your computer
instead. Give me a break. The telephone is convenient, it works,
it goes in your pocket. The mistake he's making is to think that the
world is going to be a neater place, that Internet Protocol will do
everything. I think it'll be quite the opposite. Things are getting
more diverse.
Q: Can you give me some examples of that?
A: You are going to see other protocols for things like
channelized data, where you want your own pipe or you need
more security. And you are going to see a multiplicity of devices,
not a blurring. You don't want to watch video on your cell phone.
Within five years, every new car in the world will have a
satellite antenna that lets it receive 500 radio stations and six to
eight hours of storage for saving programs. And think what your
life will be like when a TiVo box [a digital video recorder] will
be able to store 3,000 hours of video instead of 30. All of these
different devices will use various kinds of networking
technology.
Q: So what is the hottest area you are looking at now?
A: I'd say metropolitan area networking. There are lots of
companies making gigabit Ethernet equipment, and on the other
side, outfits like Global Crossing (GBLX) and Qwest (QWST)
that have huge data pipes. But people don't understand that these
two worlds don't connect very well today. It's the part of the
python where the pig is stuck. So I'm looking at a portfolio of
companies that are taking a fresh look at how to weave together
these two worlds. Companies like Mayan Networks, LuxN,
Astral Point, and Quantum Bridge. They're throwing electronics
at the problem, collapsing everything together to improve the
connection between local area networks and the backbone.
Q: What's the biggest trend you see overall in the computer
business?
A: The move from products to services. The only way people are
going to be able to make money is on the service side. The
margins in the PC business are gone; there's no value in stuffing
boards. Even Michael Dell could finally work himself out of a
job. He makes only $200 off a computer that costs its buyer
$20,000 in service and support over the life of the machine. He
has to find a way to grab the other 99% of the value of each
machine his company sells. You have to keep reinventing
yourself.
Q: Are there any other technical trends people aren't aware
of yet?
A: There is tremendous stuff going on with electronic displays.
Right now, display panels cost around $1,000. But people are
figuring out how to make them as thin and flexible as a plastic
vegetable bag. Imagine what that will mean. You will be able to
hang them on your walls just like posters. Displays will be so
cheap that packages will have their own displays.
Q: What's the biggest difference that you've noticed
between the East and West Coast high tech businesses?
A: There is a tremendous amount of diversity here. But
probably the most important difference is in the work style.
People out here are less worried about failure than they are
back East. Even the venture capitalists are different: They
use the same words, but they tend to be more conservative in
the East. I saw that when I came out to learn about startups
for Lucent. When you are farther away, it's easier to deny
what's happening here. But there is a tremendous amount
going on.
Message #15106 from Doug at Nov 25 1999 12:29PM
fumble: Carrier telecommunication is faced with an electronic bottleneck which is defintely going to be replaced by light and Optelectronics.
I appreciate you cannot replace the existing protocol zoo.
However the bandwidth of usable Optic frequency is 75 thousand gigahertz. This bandwidth can be used to accomodate the zoo. Using MPLS,IP and
fibre channel
switching it is feasible to eliminate ATM completely.
I strongly believe that LU and NT are working towards that end and have slowed down R&D on electronic switches. Are you aware of what
development work these
Companies are doing on next generation Silicon switches.
Thanks. |
