What Indian IT achieved in the last two decades, was despite its infrastructure. Yes, it helped that special tech parks isolated exporters from the infra-vacuum. And that bandwidth, power and special clearances were easier to come by for exporters – allowing even infrastructure-demanding areas such as call centers to spring up from zero.

Infrastructure decade
If this was India’s Achilles’ heel through the 20th century, tech infrastructure is the strong theme of the first decade of this new century. It took off gradually following the PM’s 1998 IT task force and its 108 recommendations, two-thirds of which finally saw the light of day – a good record. And a series of policy changes, and the 2001 Convergence Act which spoke little of convergence but did replace the cobwebby 1885 telegraph act, with its separate rules for voice and data. And the convergence of the telecom and IT ministries in 2001.

There’s a telecom revolution in the air. And on the ground. Fiber criss-crosses India’s cities and countryside. Kolkata gets abundant bandwidth from Mumbai fiber landings, cities are connected by terrestrial networks, and fiber brings broadband to the kerbside for offices and homes. All that it needed was for the government to step out of the way. Pioneering this revolution has been a group of private players led by Bharti, laying fiber, offering cheaper and cheaper call rates, and pushing long distance national and international phone call costs down sharply.

Cheap and always-on
Cheap bandwidth and always-on Internet access for homes in all cities and most towns will start off in 2003 through DSL—over cable-TV cable and plain old PSTN copper pairs. Yes, it’s sort of there in a few metros, but nearly invisibly. The key players will be BSNL, Bharti, Reliance and MTNL—mostly offering DSL over their phone wiring into homes.

The competitors for access will be the surviving conventional ISPs such as Satyam and VSNL, using both satellite and terrestrial radio and fiber for backbone links, and offering dial-up access and carrying IP over Ethernet on the last mile—into offices and home complexes.

Convergence: Just the medium…
The most used and abused word of the tech era will shape the next big inflexion in infrastructure evolution in the years ahead. That’s with the striking down of the key barrier to convergence: the forced and continued separation of voice and data.

When the Convergence Act, part 2 comes in—more likely in 2004 than next year—it would drop the artificial barrier between voice and data. Corporates would pay for bandwidth and run whatever they needed on it—voice, data, video. One strand of fiber would carry data as well as hundreds of channels of voice, flexibly multiplexed as per need. IP telephony would be allowed, and not just on closed corporate networks. That’s when you’d be able to have a common LAN carrying voice and data in the enterprise. This convergence has happened in the world, while we’re still stuck on the policy barrier in India.

Cheap PC, competition and Linux
Two million units a year and a base of ten million sound like two healthy figures. But to really dilute a figure, just divide it by a billion—the population of India. One PC for 100 people makes for poor statistics. We’ll reach 4 per 100 by 2010, helped by falling prices but impeded by rising population, but it’s a moving target. China will have moved ahead to a likely 7 per 100.

But cheaper PCs will help. Cheap processors from AMD and Via will see prices slip to the Rs 10,000-levels. They’re not yet there. The Rs 10,000-PC from eSys, which runs a Via processor, goes up to 15k once you add a monitor and taxes, even if you choose to run Linux, which is what it ships with. But the sheer target of a near-10k pricepoint will force prices down elsewhere, and Intel and its GIDs will have to offer Celeron-based systems at comparative prices. They’re getting there.

Access over density: The cybercafe
Sam Pitroda’s "alternative vision"—that more than teledensity, which was a struggle for developing economies, access was the key—translated into a half-million or more long-distance phone booths that changed the landscape in India. With simple user-side metering and billing, hundreds of millions of Indians could call relatives and friends in others states—without any phones at home. The vision translates rather well into the cyber age.

Access to information and IT is not restricted by individual ownership and buying power, but by access—cybercafes, community access booths, PCO or phone booths powered by technologies such as Ashok Jhunjhunwala’s WLL-based CorDECT. In the rest of this decade, this will become even more crucial, as a new generation is exposed to IT and information.

Mobile India? Dial 802…
The next big thing waiting to happen (for the past decade) is the notebook PC. At just 5% of the market, it doesn’t sound like it’s worth writing about. On the other hand, that translates to one lakh laptops a year, which means some brisk competition between IBM, Compaq/HP, Toshiba, et al. But just as wireless changed telephony in India, Dataquest believes that wireless access could be the turning point for laptop usage in India. One, 802.11, which will be "served" in most top hotels by 2003, and in airports by 2004, will let executives "do their email" instead of wasting their time waiting at airports or lounges. Two, GPRS, together with combo devices such as Nokia’s 802.11-GPRS PC Card will push this further ahead. The peer effect of all this will mean more laptop purchases, going up to over 10% by end-2004—or a base of a million units by end-2005.

The hardware half-decade?
And finally, ten years after software, manufacturing could begin to catch up. With a very few pioneers—primarily Celetron and Moser-Baer, but also domestic-area players like TVS and WeP—showing the way, things could become easier as infrastructure, port facilities, airports and customs facilities and attitudes improve. On the other hand, the marginal duties protection that’s there will come down, so the only differentiator will be efficiency and market proximity leading to cost advantages.

Prasanto K Roy