Converting voice into IP will be the ultimate conversion

Dr Gururaj "Desh" Deshpande is a visionary and an entrepreneur of
the New World order. He has been instrumental in starting several hi-tech
start-ups like Coral Networks in 1988, Cascade Communications in 1991, and
Sycamore Networks in 1998, all in the US. He was in India in July this year to
announce his first Indian initiative, Tejas Networks. Voice & Data caught up
with him for an insight on convergence, technologies, and entrepreneurial
spirit.

How do you view convergence that everyone is talking about today?

One of the ways to make convergence happen is to make data pretend like voice.
You build a network for voice and then the data can move on it. That suited when
traffic had 90 per cent voice and 10 per cent data. But as data becomes the
larger portion, you have to make voice pretend as data. So you convert voice
into IP. That will be the ultimate conversion, where you convert all information
into IP packets and build large networks just to communicate IP traffic.

But obviously that conversion will take some time. You need to build the
infrastructure that facilitates that convergence.

What we are doing is championing the technologies to build the fiber-optic
backbone networks, where there would be co-existence, i.e., you can have the
current existing voice networks and on top of that you can put data. And when
voice gets converted to data, then it can be more and more of data.

Everyone–a vendor, a user, and a service provider–looks at convergence
from his own point of view. How do you see the convergence coming about?

Each will be driven by different factors. For the end/home user, it is the
question of cost. For the application provider, it is the new applications, like
the CRM packages. For the vendor, it is trying to support those two facets–you
want to build a network so that you can bring the cost down and it does not make
sense to put two different infrastructures. Vendors will have to come up with
new applications so that voice, data, and video can go together and give you a
richer experience in communications.

Several players like Lucent believe that next phase after "Internet
economy" would be the "Light economy". How do you envisage it?

I would like to call it a "heavy-duty economy". It is the bandwidth
economy and networked economy. In such an economy, we need a lot of bandwidth
and the only way to create abundant bandwidth is optics. In the past, a lot of
R&D activity was done for making the best use of whatever little bandwidth
you had. In fact, it is still happening because the world will need a lot of it.
And you can actually create the bandwidth using fiber optics. So you need to go
from ration model to abundance model.

Does that mean "bandwidth on the fly"? Does it really warrant the
need and how can that be made possible?

Look at the scene right now in India. You can get an ISP connection with modem
and DSL. It could then be wireless. All this would choke the backbone. The first
stage is to get a lot of capacity on the backbone for the ISPs. But as soon as
you do that, you need to get all kinds of new services, videoconferencing or
downloading big files. That is when you need to be able to set up
"bandwidth on the fly". Ultimately, it has to be bandwidth at a phone
call–you make a phone call and get gigabits of capacity.

Intelligent optical networking will help in resolving the challenges of
bandwidth. The basic thing in optical networking is that you can send multiple
lasers from a fiber. You can make one fiber look like multiple fibers. However,
just taking a lot of bandwidth is not enough. That kind of network was okay when
we had only phone calls. The New Networks will have to support a lot of new data
and unpredictable traffic pattern. The new intelligent optical networks use
optical routing of signal, in addition to optical transmission, and deliver
network services. These networks have scalability and are cost-effective.

Continued...

Where exactly do the intelligent optics fit in the public networks?

Divide the market for optic networks into three segments–the long-haul market
(transport and switching), the metro market, and the access market. Intelligent
networks will fit in all the three.

What would be the rational model for building the fiber-optic backbone
networks?

The old transport technology was like having a mainframe in the organization for
computing. It took lot of time, lot of planning, implementation, inflation, etc.
The typical business model was you get the Right of Way (RoW), put the fiber in
the ground, and give big purchase orders either to Nortel or Lucent. They go and
build the network for four-five years and then use it for 30 years.

The new technology is like buying a PC. You can bring the networks up within
days. You do not buy PCs for the next 30 years as the technology will get
obsolete. Once you have the RoW, you lay the fiber. And once you have the fiber
you can put something here and some there and, bang, the network is up. You can
bring the network up rapidly, without building the whole network at one go. You
can build it in steps according to usage and incrementally expand your network.
The carriers need to understand this and pick the best route. The new optical
intelligent networking equipment that goes into the New Network enable them to
do just that.

Fiber in India is not used optimally. What are the broad measures that
need to be taken to prop up the infrastructure quickly in India?

The only one to have access to the available fiber is DoT. So DoT needs to speed
up. But being a monopoly, it can move at its own pace. Now others are laying
fiber and they have expectations to meet. And their pace should not be such that
they own the fiber in the next three years, design for two years, and be up for
another five years. They need to do things in a record time. That is the new
mindset that we need in this country. There needs to be that level of pride.
There should be a determination in the carriers to beat every record. There
should be a feeling that "I am going to do it faster than he did it".
We need to get the business moving.

And how do you get the business moving?

Through competition. You deregulate and then the opportunities are absolutely
clear. If I can swing a fiber and light it, let us say between Bangalore and
Pune, I can sell all day. So once they deregulate, people will see
opportunities. When one guy comes in and makes lot of money, the second one
jumps in, then the third, and the fourth, and then the next…and there is
competition. That’s how we need to get going.

So, the most important thing is deregulating. World over, it is a proved
phenomenon that the more you deregulate, the more competitive the market is, and
the better is technology injection into the market place.

Continued...

Is small beautiful or does size matter? Most companies start small. At
that point of time they attribute flexibility and start-up mentality as the key
attributes for growth and wealth creation. Then as they grow, they consolidate
through M&As. How well does that work?

It is not small which is beautiful, but entrepreneurial that is. So a small
entrepreneurial group is lot more powerful than a large bureaucratic group. But
if it is a large entrepreneurial group, it is the ultimate. That is what we are
doing at Sycamore. Tejas is an entrepreneurial group and by associating Sycamore
with Tejas, it is going to be powerful. We acquired another entrepreneurial
group called Sirocco Systems. It is not acquisitions, but a question of how you
continue to work.

The reason why a lot of these acquisitions don’t work is that you kill all
the creativity and do not allow the entrepreneurial spark to ignite. We want to
do business with Tejas not because we want them to do the things that we want,
but because they have ideas and a mission. You need a mission that is ambitious
enough to accommodate a lot of those entrepreneurial activities. Sycamore will
continue to take this entrepreneurial spirit to pursue reinvention, creation,
and innovation of newer technologies. So M&As are not bad, but it is only
that they have to be done the right way. You have to do acquisitions out of
strength, not weakness.

The entrepreneurial spirit has caught up in India. Are we on the right
track and what will buoy the spirit?

In India, and as Indians, people are very entrepreneurial. The entrepreneurial
spirit can be easily related and explained from the cable TV example. But what
we need next is the opportunity to play the big game. I mean, you want to be
entrepreneurial but you do not want to be a beedi shop–good for a little
corner.

Dominating the market in a small place is not sufficient. You need to
dominate the global market. To be able to do that, you need connections–a
window to the rest of the world. You cannot create a global leader siting in
Bangalore because the market in India is not leading-edge enough. In the US, it
is easy because if you win the market there, then you automatically get the
opportunity to capture the rest of the markets.

The association of Tejas and Sycamore is what you need to give that
opportunity. This association is first of its kind, and when it works, we hope
there will be hundreds of companies in such associations operating in all
segments of the market. But it is very important that you have access to the
most competitive and most advanced market of the world because you have to win
there.

Further, it needs to be realized that because of the competitive nature,
there would be failures. Indians do not realize that "failure" is
essential to being entrepreneurial. You fail, admit you failed, learn from it,
and move on. You fail not because you did not work hard, but the model was not
right.

The fiber-optics market would see a mushrooming of many players. What would
be the trend and the expected numbers?

There are some players who get the RoW and lay the fiber, and some who light
their own fiber. Then there are those who sell the dark fiber and they light the
network for others. So the faster the market grows, the more will be the
start-ups. In the US, every week there are five start-ups. There is a lot of
activity going on all over the place.

The next thing will be consolidation. Once the consolidation happens, a lot
of technology comes over–again disruptive–which is 10 times cheaper. Then
lots of start-ups get in. Therefore, consolidation and start-ups will continue
to happen for a long time to come.

Worldwide, there would be 50-100 players for a long time. In terms of
start-ups, there are thousands. And that’s what needs to happen in India. You
need someone to say, "I would like to dominate between Bangalore and Hubli"
and get the fiber up and get going. You get lot of capacity and build your
business, and then build a large network. But you need to get that
entrepreneurial activity going. In revenues, the size would be in the range of
$40 billion by 2004-05.

This article was originally featured in the September 2000 issue of Voice & Data.