The XO-1 is the first shipping design to emerge from the One Laptop Per Child project, an interesting initiative blending laudable objectives with an effort to redefine the PC cost structure. The OLPC organization aims to provide the XO laptops to children of developing nations, in an effort to democratize technology for the long-term global good.
The XO-1 is designed to be rugged and low in cost, utilizing open-source software (for now) and solid-state memory instead of a traditional hard drive. Nevertheless, the design offers 802.11 wireless connectivity and features a dual-mode display; users can turn off the backlight to attain a sunlight- readable, reflective-monochrome- type screen.
|For a full archive of articles and related On-Demand seminars, click here|
A VGA Web camera, internal microphone, dual speakers, QWERTY rubber keyboard, resistive/capacitive touchpad, three USB ports, external microphone and headphone jack are other traditional PC features. Adding cost, but increasing usability, the chassis of the XO is designed with a pivoting, reversible display that may be used in a tablet-PC or e-book format.
Made in Taiwan
While OLPC drove the design and specification of the box, Taiwanese OEM Quanta is behind the detailed implementation and manufacturing. In final form, the XO-1 places the display and electronics in the swivel upper half, while the battery, keyboard, combo resistive/capacitive touchpad and little else live below. The arrangement runs a bit counter to the mainstream market, which usually puts electronics, keyboard and battery in the lower notebook half, with only the LCD panel above.
The dual-mode LCD–supplied by AUO–is a curious bit of technology. Despite much talk of fancy diffraction gratings and other trickery to implement a combination color-backlit and monochrome-reflective display, we couldn't detect anything other than a fairly standard pixel stackup and construction vs. a standard transflective-LCD construction. The RGB pixel filter arrangement is based on diagonal striping instead of the normal adjacent RGB triad, and thus requires some special addressing and dithering. Otherwise, we saw little evidence of a radical approach to the LCD.
White LEDs are used for backlighting in transmissive mode. Fluorescent tubes have been the norm in backlighting for cost-sensitive panels of the size used in the OLPC, but perhaps power-efficiency considerations won out over cost here.
Power savings a priority
Certainly, reduced power consumption for the OLPC drove the reflective-monochrome viewing mode. The guiding principle is that developing countries would often lack the energy sources to provide for frequent recharge or plugged-in operation. The original OLPC prototype design even sported a hand crank for recharge, though this solution faded as the product evolved. Presumably, the true power budget began to swamp what a human-power generator could achieve with the small muscle groups of the arm.
In a further effort to lower power draw, the Himax HX8837 LCD timing/controller chip allows the CPU to sleep without loss of the displayed image, for power savings. An image buffer for this device comes by way of an ESMT M12S16161A 2-Mbyte mobile SDRAM. In full-tilt operation, we observed about a 6- to 7-watt power consumption. The OLPC's roughly 20-watt-hour LiFePO4 battery would thus supply up to three hours of operation in most-active use. Preliminary reports of real-world battery life peg the 10- to 12-hour run-time claim as far too optimistic, and our data would tend to agree.
The compute electronics of the XO are centered on the AMD Geode LX700 and AMD CS5536 companion device for I/O and bridge functions. The low-power chip set and solid-state memory contribute to the system's ability to operate without need for a fan. A single 1-Gbyte package of NAND flash–consisting of two SLC NAND dice at 512 Mbytes each from Hynix (the HY27UG088G5M device)–provide the operational equivalent of a hard drive, though clearly falling well short of the normal expected nonvolatile storage of the normal PC.
Operating memory is provided by Hynix as well, in the form of four single-die packages of 64 Mbytes each (HY5DU121622CTP), for a total of 256 Mbytes–again, a bit thin vs. the mainstream but adequate for the task. The audio features are handled with an Analog Devices audio codec(AD1888) and stereo amplifier (SSM2302).
The XO-1 creates its own mesh network out of the box. Each machine is a full-time wireless router, such that multiple Wi-Fi-capable OLPCs are made aware of one another and can route information across and through the mesh. Flip-up Wi-Fi antenna “ears” on the edge of the display do double duty as latches, holding the top and bottom halves together in the closed position.
Connectivity is provided by a Marvell solution on its own substrate: a 802.11 transceiver (88W8015) and 802.11 system-on-chip (88W8388), with a 16-kbyte E2 PROM from STMicroelectronics (M95128). Marvell also provides the SD memory and camera controller solution in the form of the 88ALP01.
User input is accomplished with a membrane keyboard comprising two PET films and one elastomer keypad. As in most keyboards, elastomer keys cause contact to be made between two PTF- silver/PET films, completing the circuit for an individual key. An NEC 8-bit microcontroller (UPD78F0511) handles the interface between the keyboard and the keyboard controller from ENE (KB3700).
STMicroelectronics' touch controller provides the interface chip between the Alps-manufactured resistive-and-capacitive touchpad. The capacitive touch film uses a set of underlying technologies very similar to that for the films referenced for the QWERTY keyboard–all fairly simple and inexpensive, at least in appearance.
Despite finally shipping product, the OLPC effort hasn't gone flawlessly. Delays and cost overruns, along with a host of political hot potatoes, have given the project some black eyes. The current price tag for the laptop has ballooned from the original $100 objective to almost twice that, and our cost estimates–available in a more detailed report from Portelligent–show the reasons why.
In short, displays, batteries, CPUs, fancy enclosures and the rest all add up to a bill of materials that sits uncomfortably far from the $100 target. OLPC may get there, but it has a ways to go.
Can more widely available connected computing help bring countries from “developing” to “developed”? Only time will tell, but if nothing else, the OLPC effort helps change the way the industry thinks about how expensive a PC really needs to be. Surely there will be beneficiaries along the way, direct and indirect, even if the project never reaches its full set of philanthropic objectives.
An acceleration of cost-reduction pressures for the larger notebook industry may be OLPC's biggest impact, at least outside the developing nations targeted by the design.
David Carey is president of Portelligent (www. teardown.com ), a TechInsights company. The Austin, Texas, group produces teardown reports and related industry research on wireless, mobile and personal electronics.