A case for custom silicon in enabling low-cost information technology for developing regions

Abstract

Information and communications technology has the potential for deep social impact in developing regions but today's typical ICT devices -laptops, mobile phones, and similar devices -are often still too expensive for many scenarios. In this paper, we argue that custom integrated circuits can enable a new tier of low-cost information access devices with a price point that will make them widely accessible. And, with control over the silicon, these systems can economically address many other challenges. To evaluate our claim, we focus on a deceptively simple problem -low-cost information access for illiterate populations through audio recordingsand show how custom silicon allows us to reduce cost, lower power, leverage conventional infrastructure in unconventional ways, and optimize the interface for usability. In particular, we show how a rural audio computer can be designed around just three chips, use an inexpensive capacitive touch interface, employ inductive communications for peer-to-peer data transfer, and employ content download over GSM voice and FM broadcast as two wide area options. The resulting design point -enabled by aggressive silicon integration -affords a device that can be built for $7.77, a third of the cost achievable using commercial off-the-shelf components.

Raabta

“…Cost: Roughly half of the population of the earth lives on less than $2 a day [10]. Even basic electricity is a luxury for this cross-section.…”

“…Power: Most of the developing world is either without power [10] or experiences intermittent and bad quality electricity [5,27]. Therefore, low-power consumption is also a first-class goal.…”

“…Table 1 dissects the cost of each component. As opposed to desktops and laptops, which cost several hundred dollars, this entire setup costs $92, which we believe can be further reduced through the use of custom Silicon [10]. Although our original goal was to make the end-host accessible to every household within the rural community, in its current form the village kiosk model might be more applicable [5].…”

“…In particular, the codesign of the system, the use of multi-layering encoding, and the dynamic switching between FEC and retransmission is a direct consequence of the insights they provided into the internals of Skype and Google+. Moreover, Raabta shares goals with the rich-set of work dedicated to providing Internet connectivity to the developing world [17,4,18,20,21,14,10]. But in contrast to Raabta, all of these initiatives require both the deployment of new infrastructure as well as Internet backbone connectivity.…”