If we are to get to the next billion users, the challenge before us is as follows: create a computing solution at a significantly lower price point, which is also more easily manageable. At the same time, it should be able to support much of the applications base that has already been developed, even as it needs to encourage new applications development for the next set of users. The need is for a whole solution, not an assembly of parts. The need is for a disruptive innovation in the PC industry.
There are four key ideas to bring about this revolution:
Server-centric computing: The benefit of having fast, reliable networks is that it does not matter where the storage and processing is happening. So, why not centralise it? Leverage Moores Law on the server. The server can be on the enterprise network (in the case of the SME) or on the operator network (in the case of the home users). In both cases, the networks are either LANs or extended LANs with connectivity from 10-100 Mbps. Once we assume that this connectivity is going to be there, it becomes easy to think of centralizing processing and storage on servers.
USD 100 virtual PCs: Server-centric computing now simplifies the clients dramatically all they need to be are graphical terminals which connect to the servers, and are maintenance-free and upgrade-free, for life. It should be possible to get the cost of these terminals to between USD 100-150. The actual computer needs to be no more than USD 50 (a thin client) all that is needed is a processor, which can run a software like vnc (virtual network computing) that drives the local interfaces. The five connectors that need to be available are for keyboard, mouse, display, LAN and power. The biggest cost will be the display. Refurbished monitors will cost USD 50 while new ones will cost USD 100.
Open-source software: The underlying software needs to be based on Linux, with other open-source software used for the various applications. Linux is more than good enough, and there are thousands of applications available. Emerging markets need to shift from Microsoft Windows and Office even though the software today may be theoretically cheap because it can be pirated, these markets are damaging their own long-term futures by reliance on robbery rather than legal alternatives. Piracy stunts the growth of their own domestic software industry.
Remote support: Once the endpoints (desktops) are simplified, support only needs to be provided at two levels: to end-users for applications, and for the server. Both of these can be offered remotely in very much the same way companies in India are offering outsourced support to global organisations. This gets around the problem of cost-effectively supporting smaller technology set-ups which will be the case in the events of homes and SMEs.
Tomorrow: The Economics and Ecosystem
TECH TALK The Next Billion+T