Demo To Production, Intel's 80-Core Chip Faces Hurdles
The experimental chip can perform specific calculations, but it isn't yet capable of performing a wide variety of functions, analysts say.
Intel's experimental 80-core microprocessor is a marvel in processing power and low energy consumption, but how quickly a chip of this caliber enters the market will depend on how fast Intel can cross a number of hurdles, experts say.
The chipmaker unveiled the Teraflop Research Chip -- codenamed Polaris -- on Sunday. Intel plans to demonstrate the fingertip-size chip with the power of a supercomputer at the International Solid-State Circuits Conference this week in San Francisco.
The Polaris chip can crunch a trillion mathematical calculations a second, a feat that 11 years ago would have required a 2,500-square-foot supercomputer. Polaris's size and the fact that it uses about the same power as today's two-core processors mean that it could someday provide unprecedented processing power within a home computer.
Such muscle would be useful, for example, in developing video games that look as real as TV shows, or in providing imaging recognition, which means a computer could recognize photos and catalog them appropriately. But beyond gaming and specialty applications, the need for a desktop supercomputer in the home is not yet clear. "That's a little bit nebulous," Martin Reynolds, analyst for Gartner, says.
But assuming that the technology industry finds a use for Polaris in the five to eight years it will take Intel to move from experiment to production, there are still a number of challenges that will have to be met before the chip enters the consumer or business markets.
First off, Intel's experimental chip can perform specific calculations, but it isn't yet capable of performing the wide variety of functions handled by today's general-purpose two-core and four-core microprocessors, says Jim McGregor, analyst for In-Stat. "This is only a test chip; it's not something that will be produced [for the market]."
Then there's the dramatic change in architecture. Polaris divvies a computational task among all 80 cores, which each process a piece of the problem, and delivers the solution as a whole. While the process is similar in today's multicore chips, Intel developed the algorithms necessary to divide tasks among the dozens of cores in this processor.
This different architecture will require more intelligence in a computer's operating system, as well as in the software that runs on top of the OS, experts say. Then there are the new application development tools that companies such as Microsoft, Borland, and IBM will need to develop. Developing multithreaded software that takes advantage of multicore chips is a task that many enterprise software companies are just now beginning to undertake.
Developers often write code in building applications for today's simpler chips, Reynolds says. With a microprocessor as complex as Polaris, code won't be written, but more likely generated from models. McGregor agrees, saying, "It's a whole different ballgame."
"Think of the human brain. We can't think of anything simultaneously, and yet we're building a chip that can," McGregor says.
Nevertheless, Intel's timetable is certainly possible, given that it has already developed the technology for large-scale parallel processing. But the microprocessors that hit the market in five years or so probably won't have 80 cores. "It's a very feasible timetable, but not at 80 cores," McGregor says. "You could, however, see chips jump from four to 20 or 40."
The Business of Going DigitalDigital business isn't about changing code; it's about changing what legacy sales, distribution, customer service, and product groups do in the new digital age. It's about bringing big data analytics, mobile, social, marketing automation, cloud computing, and the app economy together to launch new products and services. We're seeing new titles in this digital revolution, new responsibilities, new business models, and major shifts in technology spending.
Join InformationWeek’s Lorna Garey and Mike Healey, president of Yeoman Technology Group, an engineering and research firm focused on maximizing technology investments, to discuss the right way to go digital.