Software’s Bandwidth Barrier

In
a recent issue, Fortune magazine dedicated a special section to Indian IT czars
in Silicon Valley (most of whom belong to the software world). Fortune spoke in
glowing terms about these icons beginning to dominate this Western IT Mecca with
their bold, brilliant and buoyant ventures.

We couldn’t help but feel proud of the massive strides
India has made–not just in terms of individual successes, but as a nation–in
the software and IT markets across the globe.

The Indian software industry has had humble beginnings–our
software engineers were earlier treated as code developers and Indian software
companies, sweatshops supplying cheap intellectual capital. Over the years the
industry has shaken off this stigma–emerging instead as a powerhouse of
"highly skilled" manpower with a feel of every single cutting edge
technology and platform. Indian software companies are for instance supplying
everything from animation software to WAP solutions, from browsers used on new
generation wireless phones to e-commerce web sites.

There’s bigger growth post-Y2K

The software industry in India grossed US $5.7 billion of
annual revenue in the 1999-00 fiscal with almost $4 billion coming out of
software exports. Once again, the over 55% growth was reason enough to shut up
critics, who had proclaimed that after Y2K, the Indian software export segment
would slow down. On the contrary, the JFM quarter of year 2000 contributed
almost 32% of the software revenue in fiscal 1999-00!

The fiscal year 1999-00 also witnessed the maximum demand for
software professionals from India from other countries. The estimate of demand
was anywhere from 200,000 professionals to half a million. I am sure that India
will take on this challenge and meet such demands.

One
of the most noticeable highlights of 1999-00 was that 185 of Fortune 500
companies outsourced their software requirements to India. During the year,
companies in India continued to get SEI/CMM Level 5 Quality certifications in
abundance and by August 2000, out of a total of 32 companies worldwide with SEI/CMM
Level 5 certification, 16 were located in India–that’s 50%!

In the last 18 months, five Indian infotech companies have
listed their ADRs in the US. Infosys, Satyam Infoway, Rediff.com, Silverline and
VSNL have led the way, with 20 more in the pipeline, waiting to list at Nasdaq
or NYSE over the next 15 months. This has made US investors further recognize
the strengths of the Indian software and services industry, and in turn making
it easier for many Indian software companies to export to the US. In 1999-00,
almost 60% of India’s software was to the US.

India also continued to increase its software exports to
Europe. During 1999-00, out of a total of $4 billion exports, close to $1
billion was exported to Europe.

The big news of 1999-00 was the growth of market cap of this
industry. The market cap of the software industry went up from $4 billion in
January 1999 to almost $104 billion in February 2000 and stabilized at about $65
billion by 31 August 2000, a distinction which made the industry a "best of
wealth industry".

The domestic market continued to gain and went up from $1.2
billion to almost $1.75 billion. The segments of the government, banking and
SOHO were the real highlights. Piracy figures came down from 63% to 59%.
E-governance, however, was mostly just talked about–but is yet to take off in
a big way.

However, in all this euphoria, the situation of telecom
infrastructure and bandwidth remained in a state of flux, and the cost of
international leased lines remained high.

Bandwidth: the last frontier

To become a software superpower, India needs to provide,
apart from other things, adequate, affordable and reliable bandwidth.

In simple words, India requires quick implementation of a
robust telecom infrastructure. At the same time, we also want sane policies so
that we do not stifle technology. Take voice over IP: India is probably the only
country actively attempting to ban this technology, whereas voice over packets
is the direction world telecom is heading toward. The big Western telecom tech
giants have already begun moving from circuit-switched to packet-switched
long-distance voice networks.

Voice over IP must be allowed in India. Perhaps in a phased
manner, and while taking into account the concerns of the existing players.

But first let’s address the issue of bandwidth itself.

With a $4.1 trillion market potential, e-commerce has become
the engine of growth in the new century. India has done well in software–but
this is only the tip of the iceberg. The future is in leveraging the IT skills
in the country for e-commerce, IT-enabled services, IT services, et al. The
skills have to deployed in a certain environment and with a minimum level of
guaranteed infrastructure, particularly with interactive applications.

So far so good. But all this requires and depends on a very
strong telecom infrastructure. Specifically, the availability of adequate and
affordable bandwidth.

In this present Internet-based economy, per capita bandwidth
utilization is an important indicator of the economy.

Net age measures....>>>>>

Net age measures

Based on Nasscom’s rough studies and projection, I propose
that in the next five years, we aim to provide at least 2Mbps bandwidth to every
citizen of India.

Currently, only 325Mbps of international Internet gateway
bandwidth is available from India. The demand, meanwhile, with some 1.5 million
Internet connections, has already touched 5Gbps, and is expected to swell to
10Gbps by year-end. In a comparative situation, China already has an
International bandwidth of 55Gbps, the UK has 120Gbps, and tiny Japan has
160Gbps.

The present international bandwidth to the tune of 10Gbps is
available to VSNL through the submarine cable FLAG (Fiber Optic Link Across the
Globe), while another 20Gbps is available from SMW3 (Sea-Me-We-3). This is used
both for voice and data. Another 20Gbps is going to be commissioned by SAFE at
Cochin.

The problem here is that VSNL takes almost 60-100 days to
provide 2Mbps of bandwidth, and that too, at a very stiff cost.

The other–possibly even bigger–problem in India is the
backbone. The bandwidth available to Internet traffic on the domestic routes is
a mere 8 to 34Mbps. Thus, there is a complete mismatch between demand and supply
of Internet capacity in India.

A software or service company which takes a 2Mbps Internet
gateway connection at one location in India cannot easily distribute that
bandwidth to its other location–it is unlikely to find available capacity on
the domestic routes, thus forcing it to take another expensive gateway. This is
the same situation ISPs are contending with.

In addition, 2.4Gbps of national bandwidth is being used in
India for voice communication on major trunk routes. The bandwidth on
lower-traffic domestic trunk routes is about 560Mbps.

"But where’s the demand...?"

It has to be understood that in the Internet business, it is
the increase in supply of bandwidth which itself creates higher demand. This has
been already proven in the USA and Europe. We need to change our mindset in
India, and not wait for demand and pending application backlogs to grow before
we provide bandwidth.

According to Nasscom’s Survey, the International and
national bandwidth requirement in India by 2005 is expected to reach 300Gbps,
from the current demand of 10Gbps. In the future, the main demand of Internet
bandwidth would come from the fields of B2B and B2C e-commerce, communications,
entertainment, residential video services, software exports, education,
IT-enabled services, and banking and finance.

Let’s see, in simple terms, what kind of bandwidth is
required for next-generation applications.

For example, we require 64Kbps for basic animation services;
200Kbps for distance education; 10Mbps to 1Gbps for virtual reality, and so on.

Operation Bandwidth

Inspired by a Voice & Data debate on bandwidth in March
and Dataquest’s April 15 editorial on the subject, Nasscom launched
"Operation Bandwidth" in May.

Operation Bandwidth asked for 300Gbps by 2005 and made a
number of recommendations, giving their impact (see report in this issue).

And it was taken seriously by the government. Among the
initiatives that the government launched in response:

• It formed a high-powered bandwidth committee at the
  Department of Telecom, to: (a) Forecast short- and long-term bandwidth
  demand; (b) Assess domestic and international bandwidth availability for
  India, and the steps needed to fill the gaps; (c) Streamline procedures for
  making available bandwidth as well as providing network/leased circuits; and
  (d) Identify the causes of delay and recommend policy and process changes to
  ensure prompt availability of bandwidth and network/leased circuits.

• The prime minister announced the opening up of national
  long-distance telephony (interstate STD calls) service to private operators,
  to encourage competition.

• The PM also announced the complete de-monopolization of
  VSNL in the area of international fiber-optic capacity, licensing and usage.
  With respect to international connectivity, direct contracts between
  submarine OFC operators and ISPs have been allowed. Thus, private ISPs will
  be allowed to set up their own landing stations anywhere in India singly or
  jointly, in collaboration with international undersea bandwidth carriers. (VSNL
  has been advised to renegotiate its contract with FLAG to allow FLAG’s
  unused capacity also to be used here.)

• VSNL announced "bandwidth on demand" and
  reduced the days required for connectivity from 153 to 60.

Many more steps need to be taken to improve the bandwidth
conditions, and both the public and the private sector need to invest.

The Infotech impact

The year 2000-01 promises to bring even more gains than last
year. Software exports may gross as high as $6.3 billion, which means a 60%
growth. The star sectors may be areas of e-commerce solutions, IT-enabled
services, software engineering, systems-on-a-chip, ASP (application services)
and interactive architectures. In addition, call centers, animation and content
development may bring the icing on the cake.

That will probably be the biggest challenge in the year
ahead.

But the software export results during 1999-00 and the
prospects of 2000-01 clearly show that the Nasscom-McKinsey study projections of
$87 billion of annual revenues, with almost $50 billion from software exports by
2008, can be achieved even earlier–if the bandwidth and the knowledge workers
situation both improve.

Dewang Mehta

is president of Nasscom, the National Association of Software and Service
Companies