Split Chip Architecture Could Lead to Smarter Smart Phones
U.K. chip manufacturer ARM last week unveiled a chip architecture that could simultaneously bring down the cost of smart phones while extending their battery life.
The architecture combines an ultra-efficient new processor with a high-performance version on the same chip. A smart phone will be able to switch between the two depending on the task at hand, says Nandan Nayampally, ARM’s director of CPU product marketing. Nayampally adds that operating system software will need to be rewritten to take advantage of this dual-brain feature.
ARM licenses processor and chipset designs to other companies, which then manufacture the chips. The company has specialized in small, low-cost chips for many years, a strategy that has put it in an ideal position to capitalize on the shift toward mobile computing. More than 90 percent of all smart phones use chips based on ARM designs.
In the past couple of years, Intel has sought to catch up, by developing a line of chips, known as Atom, for mobile devices; but these have not so far matched the performance or efficiency of ARM-based hardware.
For high-performance applications such as Web browsing, navigation, or gaming, software built into the chipset will use the powerful Cortex-A15 MPCore processor. For less-demanding background tasks, such as voice calls or text messaging, it switches to its new low-power Cortex-A7 processor.
“Smart phones already have dual-core processing today,” says Nayampally, referring to chips such as ARM’s Cortex-A9, which contains two separate central processing units. By dividing up data and allowing it to be processed in parallel, programmers can wring more performance out of dual or multicore chips. ARM’s new architecture extends this model of multiprocessing to processors of different size and performance.
Using two different processors is at odds with most practices, where the mantra is to pile in as much power as possible into one processor, says Mark Zwolinski, of the electronic systems design group at the University of Southampton. “It’s quite a neat idea. It’s like all great ideas—it’s blindingly obvious in hindsight,” he says.
The A7 processor is more efficient, primarily because it is physically much smaller. At just 0.5 millimeters square, it is one-fifth the size of typical smart-phone processors, and it uses 20 percent of the power.
Normally, this would reduce the speed of the processor. But ARM used new techniques, including a 28-nanometer chip fabrication process instead of a 45-nanometer one. As a result, the A7 is still powerful enough to run a basic smart phone. Because the chips are smaller, and a larger number can be cut from a silicon wafer, the price of these processors should be low enough to make smart phones that will sell for less than $100, which could make smart phones affordable in developing countries.
“Even $200 handsets today can run browsers and most games that the $500 high-end handset runs,” says Nayampally. The cheaper phones just might be a little slower and less responsive. But ARM says Cortex-A7 will match the performance of many existing smart phones. The Cortex-A15 will provide a level of performance well beyond what is available today.
“What’s not clear is how much of this power management is automatic and built into the hardware, and how much needs to be built into the software operating system,” says Zwolinski. Operating systems such as Google’s Android and Apple’s iOS will need to be modified to take advantage of these power savings. “So by itself, it is not a game-changer,” he says.
The new processors are now available to customers including Texas Instruments, Samsung, and Apple for testing. They should appear in devices by 2014, says Nayampally.
Even without the new technology, it will be tough for others to compete with ARM. “The company was the first to develop low-power CPU cores for licensing,” says Linley Gwennap, principle analyst at the Linley Group. “Nokia chose ARM for its early cell phones, and other mobile companies followed suit. Today, all software for mobile phones is developed on ARM, making it difficult for other instruction sets to break into that market. With more than 1.6 billion cell phones shipping per year, the volume of this market swamps that of PCs, servers, and most other processor applications.”
Geoffrey Hinton tells us why he’s now scared of the tech he helped build
“I have suddenly switched my views on whether these things are going to be more intelligent than us.”
ChatGPT is going to change education, not destroy it
The narrative around cheating students doesn’t tell the whole story. Meet the teachers who think generative AI could actually make learning better.
Meet the people who use Notion to plan their whole lives
The workplace tool’s appeal extends far beyond organizing work projects. Many users find it’s just as useful for managing their free time.
Learning to code isn’t enough
Historically, learn-to-code efforts have provided opportunities for the few, but new efforts are aiming to be inclusive.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.