India Unveils $35 Computer

Filed 24 July 2010 — New Delhi

India has unveiled a touch-screen computing device intended for students that it says can be built for about thirty-five dollars. The higher education minister presented the prototype this week and said his department had opened discussions with global manufacturers about mass production, describing the figure as covering the motherboard, processor, connectivity, memory and display together.

The claim drew immediate scepticism and immediate enthusiasm, and both reactions are defensible depending on which question is being asked.

What thirty-five dollars can plausibly buy

The number is a bill of materials at volume, not a retail price. At the scale India is contemplating — tens of millions of units — component costs behave very differently from the way they do in a consumer market. A low-power processor of the class used in phones, a modest resistive touch panel, a few gigabytes of flash and a wireless radio can be assembled cheaply if the design accepts constraints that a commercial product would not.

What the figure conventionally excludes is assembly labour, packaging, distribution, warranty, software development and support. Those costs are real and, in an education deployment, the last two dominate everything else over a device's life.

The lesson of the previous decade

Low-cost educational computing has an instructive history. The most famous earlier attempt promised a hundred-dollar laptop, shipped at closer to double that, and discovered that the difficult part was never the hardware. It was teacher training, curriculum integration, electricity in the classroom, repair when a screen cracked, and a plan for what happens in year three.

Devices that arrived without those things ended up in cupboards. Devices that arrived with them changed how lessons were taught. The differentiating variable was the support system, and support systems are the one component whose cost does not fall with volume.

Why India is a different case

Three factors distinguish this attempt. India has a domestic engineering base capable of designing the device rather than importing someone else's. It has a scale of demand that makes bespoke component orders economic. And it has a specific, articulated purpose: connecting students in higher education to a national network of digital course material, rather than distributing general-purpose computers and hoping.

A device designed around one job — display course content, take input, hold a charge, survive a school bag — can shed cost that a general-purpose machine cannot.

The honest caveats

A prototype demonstrated at a press conference is a long way from a production line. Announced price points for state-sponsored hardware have a strong tendency to drift upward between demonstration and delivery, and the gap between a working sample and a manufacturable design is where most such projects fail. The reasonable position is to treat the figure as a design target rather than a specification.

The parts nobody costs

Assume for a moment the hardware figure holds. The expenditure that follows is where these programmes are decided. A device deployed into a school system needs content in the languages of instruction, teachers trained to use it, a repair path for a cracked screen that does not involve shipping the unit across a subcontinent, electricity to charge it, and a network to reach it.

Each of those is a recurring cost, and none of them falls with manufacturing volume. Over a three-year life, the support system routinely exceeds the hardware bill — which is why the interesting number in any announcement of this kind is not the device price but the ratio of the total programme budget to the number of units.

Designing for the constraint

The engineering interest lies in what the design gives up. A device intended to display course material and accept input does not need a high-performance processor, a fast graphics path, a camera or a large amount of storage. Removing all of that removes cost, but more importantly removes power draw, which shrinks the battery — and the battery is often the single most expensive component in a portable device.

Designing to a constraint rather than to a feature list is unusual in consumer electronics and entirely normal in equipment built for a specified job. It is also the only route to the price point being claimed.

What to watch next

Whether a manufacturing partner signs at anything near the quoted cost; the delivered unit price of the first real tranche; and whether teacher training and repair logistics are funded alongside the hardware. National education policy documents are published by India's Ministry of Education, and international connectivity and digital-inclusion statistics are maintained by the International Telecommunication Union.