The LAN-Grid, Operator-Grid and Net-Grid are three variations on the same basic theme: move computing from the desktop to the server. It is not a new theme at all. The first computers did exactly that. So, why are we going back to the future?
I will answer this question from the viewpoint of the four key challenges that I believe computing faces today especially from the vantage point in the worlds emerging markets. These are the ADAM challenges affordability, desirability, accessibility and manageability. I will argue that grid computing as discussed here addresses all these four challenges to also create a ubiquitous computing platform.
Affordability: By moving computing and storage to the server and leveraging plentiful and cheap bandwidth, it becomes possible to do two things: reduce the cost of the network computer (thin client) that the user needs, and also enable a telco-like utility model for billing and services. The first point may be countered with the view that the price-points of computers are falling continuously thanks to Moores Law just look at AMDs Personal Internet Communicator for $249. My response is that this is still not affordable. The network computer (including monitor) must be in the $100-150 price point, and require zero-management. The utility-like pricing model is possible today it is mostly a financing issue. But telcos (and other broadband band owners) will be reluctant to finance a device in emerging markets that needs ongoing and on-site support.
Desirability: The grid provides an excellent platform for software developers and content providers to make available their offerings on a single platform and get paid for it based on user access. In emerging markets, piracy is a huge challenge so making software available on CD/DVD is not going to help the situation. This is akin to what was the situation in China in the gaming segment a few years ago. Now, online gaming is a huge growing business. What the grid does is create a centralised platform on which software and content can be made available and billed for, much like the value-added services on mobile phones. The grid frees up developers from worrying about distribution and piracy, and this will, I believe, lead to a positive feedback of greater availability of relevant content and applications which in turn will drive demand for the grid.
Accessibility: By centralising a users data, access to it is available from anywhere where a connected network computer can be placed. Users can be authenticated by login-password or by other biometric mechanisms. Upon authentication, the users desktop is made available instantly the way it was left the previous time. The grid should also be intelligent to resize the desktop based on the size of the display device the user is accessing it from.
Manageability: Perhaps, the biggest benefit of the grid will come from creating a computing environment which simple and requires little or no user involvement in its management. The next set of users are not that savvy and have little computer training. They need the benefits and utility that computing provides. Which is exactly what the grid computing hub with network computers as the spokes environment provides.
One of the key assumptions made is the availability of broadband. Looking at the situation in India today one may conclude that this is a far-fetched idea. Not so. The problem is not lack of broadband but the lack of computing services at the right points along the broadband pipes. The last-mile between the home or office and the broadband operator actually offers multi-megabit downstream the problem is that there is no content or services at the operator end. This is where the Operator-Grid comes in. By extension, it is also possible to get fibre connectivity between operators to centralised data centres. Thus, it will be possible to get the benefit of very low bandwidth prices because the bandwidth being used is local or national and nearly-free in terms of opex.
Thus, the grid computing platform can be an excellent foundation to build the digital infrastructure for emerging markets.
Tomorrow: Developed Market Drivers
TECH TALK CommPuting Grid+T