New Interfaces, Faster Chips, More Mobility

Intel hints at more fundamental changes at this year’s IDF, but the incremental changes are both difficult and hard to discern.

Changing the interface for a computer is no simple task. It can require huge amounts of processing power for what seems like the simplest of shifts, and done well it can eat through a battery in a fraction of the time it should last.

Nevertheless, that was the future that Intel Chief Product Officer David “Dadi” Perlumutter began painting at this year’s Intel Developer Forum. Unlike past years, when performance, resolution and power consumption were in the spotlight, this year’s focus involved a more subtle shift—but one that potentially has much more far-reaching implications.

The first step of this shift is already complete. It involves touchscreens as well as keyboards, often in the same device using designs that can be converted from one to the other, or in some cases what appears to be a PC but where the screen detaches to become a tablet.

The next step is enormously more complex. It involves voice recognition and gestures, rather than relying on touch. While much has been written about these kinds of advances, being able to interface with a mobile device by waving your hand and having a device fully understand what you’re saying - particularly the context in which you’re saying it - requires an enormous amount of processing power. But the challenge is to bring this to a mobile device, where battery technology is stagnant.

“The future is to interface with the machine in a better way,” Perlmutter said, showing a somewhat primitive demonstration of a game in which a 3D camera captured his movements for a slingshot. Smoothing out the graphics and making it interactive will require far more work. As he put it, a young child can recognize more movements than a complicated computer.

These are incremental steps, though, and they’re not the kind of massive changes that longtime IDF attendees have come to expect. What slows progress is that for the mobile market, these changes have to occur within a fixed power budget. Doubling performance only matters if it can be done with the same or greater battery life.

Perlmutter hinted at some architectural changes that will help make that possible. One involves near-threshold voltage scaling, something he referred to as “versatile performance.” As he put it, “if the platform is not warm enough, you scale down.”

To get to the next steps, Intel will have add a number of architectural changes. The first will be rolled out next year with a 22nm processor, code-named Haswell, that includes its TriGate or finFET technology. That will be followed by a 14nm chip, which Intel reportedly is already testing. Intel has been working with a variety of materials, including fully depleted SOI, and it has been experimenting with various gate structures and stacking approaches. But which ones ultimately get used depend on when it becomes economically required to change its processes and manufacturing. The company may buy some time just by using bulk CMOS combined with EUV lithography and 450mm wafer technology, in which it has invested heavily over the past few months. Bigger wafers and commercially viable EUV could well pave the way for advances at the next couple of process nodes.

Within this context, Intel also has settled—at least for the moment - on four main processors - Atom, Itanium, Xeon, and Xeon Phi co-processors. In addition, it has created an SoC for mobile phones, called Medfield, for which it has signed up a handful of OEMs such as Lenovo and Orange. But given the attention being paid to Apple and Android phones these days, and the gains that rival ARM has made with Windows 8, Intel may not be getting the attention it deserves for attempting to solve some of the thorniest problems in the history of processing.

Ed Sperling is Contributing Editor for Embedded Intel® Solutions and the Editor-in-Chief of the “System Level Design” portal. Ed has received numerous awards for technical journalism.