There is no doubting the need for computers in emerging markets. A digital infrastructure can help these nations address the pain points that plague personal life and business interactions better. For example, if India needs to ensure education for the 200 million youth of the country, computers can complement teachers to help students learn better. Computers can make businesses into real-time enterprises and thus make supply chains more efficient. Computing can help governments interact better with citizens. Even entertainment can be transformed with the availability of Massputers.
I dont think Massputers will be a reality in the quantities that are needed with either AMDs $249 device, Steve Ballmers $100 PC talk, or Microsofts cheaper (and limited) versions of Windows XP. To me, these solutions dont go far enough. All they do is protect the legacy businesses of these companies even as they experiment with ideas for the emerging markets. What they should really be doing is leveraging the lack of legacy in the emerging markets by reinventing the complete computing ecosystem and thinking of not computing but CommPuting. Here is what emerging markets need to build their digital infrastructure:
Network Computers: Yes, it is the old Larry Ellison idea. Where the Oracle chief went wrong was that he focused on replacing desktops in the developed markets and not on new users in the emerging markets. He also ignored the fact that rich client applications exist in large numbers and these will not all work in a browser. The Network Computers that I am talking about [see my recent Tech Talk series] are multimedia-capable thin clients which display virtual desktops from a centralised computing platform. It is possible to make these computers for about $50-60. Add in a refurbished monitor and you get Steve Ballmers $100 PC (or my Rs 5,000 PC). This can potentially run every application already written without modification and support client-side multimedia to do voice and video. Besides, it needs zero maintenance and so can be bundled by telcos and ISPs as part of a CommPuting bundle.
The Grid as Platform: Network computers need a centralised computing platform this is where the Grid comes in. The Grid is a collection of commodity hardware running Windows or Linux in Terminal Services mode. Think of it as Google, but also offering complete computing and storage (and not just search and mail). It is possible to start thinking of the Grid for massively large public computing because the broadband networks exist the world is awash in fibre, and wireless technologies are improving rapidly in terms of available bandwidth. The problem in most emerging markets is that the content is not available at the other end of the pipes. That doesnt need to change what is needed at the other end is the computing and storage part of the computer that we have on our desktop. The Grid also becomes the platform for software vendors and content providers to make available their offerings.
Utility Pricing: What is needed is not computers on installments (people are smart enough to look at the total cost of ownership), but subscriptions which can be started and stopped on demand. In addition, the mix of small payments via pre-paid cards is what has dramatically boosted mobile phone usage in India and online gaming in China. A similar model needs to be thought of for computing. This is the equivalent of computing in sachets. To a certain extent, as Ballmer pointed out, it is already happening via Internet cafes.
Tech 7-11s: The utility pricing argument also offers a corollary: think access, not users (as Sam Pitroda did when he launched the public call offices STD/ISD booths). Just like the 7-11 grocery stores that dot neighbourhoods across many cities, the need is for neighbourbood community computing centres. These need to go beyond the Internet cafes of today which only offer a computing device and Internet access. The Tech 7-11s would serve multiple purposes: offer a microGrid for computing, a distribution point for local broadband connections, and a training area for the next users so they can make the best use of computers.
Relevant Applications and Content: Even as the network computers and the grid address the affordability, and Tech 7-11s offer accessibility, there is a need to address the desirability issue. This has to come from locally relevant content and software applications. Think of information marketplaces to bridge information asymmetries among users, think of a library of business processes for enterprises, think of broadband content offering edutainment, think of a school-in-a-box or library-in-a-box for educational institutions. Aggregating what is already out there and making it available on the Grid is a good starting point.
By putting all of these point solutions together, it should be possible to make CommPuting available for Rs 700 ($15) per user per month including the computer (at home or work), the grid usage with gigabytes of storage space, broadband connectivity, basic set of software applications from the open-source world, and support. For educational institutions, the Rs 700 price point could be split across a few students, thus not only making it more affordable and but also ensuring that each student gets plenty of computing time.
It is this wholistic approach to reinventing computing which can transform emerging markets in the next five years, and bring in the next billion users. The competition here is non-consumption, the constraint is our own imagination. This is the big thing the industry has been waiting for — computings next Kumbh Mela.
Postscript: What should Microsoft and AMD do?
Microsoft should consider making Windows XP available for $1 (Rs 45) per computer per month on the Grid. Over three years, it makes half of what it otherwise charges for Windows XP. This is $36 more than what it makes today from users in the emerging markets! This will help strengthen the developer base and give independent software vendors and content developers who already have Windows-based applications and content an opportunity to generate revenues and expand further. It can also neutralise the potential impact that open-source software can have for users.
As for AMD, I think it will find a market for its PIC but it is likely to be not among the masses in the emerging markets, but possibly in selected niches in the developed markets (the second or third computer at home, libraries, internet kiosks, schools) where the simpler $249 device will finally fulfill Larry Ellisons network computing vision. Talk of unintended consequences!