Intel: Gearing Up For An ARMs Race

My coverage earlier today proposed several possible means by which Nvidia might procure an x86 CPU core. In case a link I included in that writeup didn’t make one of the possibilities clear to you, here’s the relevant excerpt from Suzanne’s as-usual excellent coverage of the Intel-TSMC link-up on Monday:

“We believe this effort will make it easier for customers with significant design expertise to take advantage of benefits of the Intel Architecture in a manner that allows them to customize the implementation precisely to their needs,” said Paul Otellini, Intel president and CEO, in a statement.

Translation: Intel’s not only using TSMC as a foundry for its own upcoming Atom-based products, the company is also willing to entertain licensing the Atom x86 core to others. This particular and all-important aspect of the agreement received a bewildering dearth of coverage from the mainstream technical press; unsurprising, alas, but highly disappointing nonetheless.

In a general sense, Intel’s core licensing decision further ratchets up the versus-ARM competition that the company’s long been instigating. Granted, the bulk of today’s Atom business resides in the realm of netbooks, nettops and small-form-factor single-board computers. But Intel’s aspirations are far broader, extending to the cellular phone (specifically smartphone) market segment currently dominated by ARM’s numerous licensees, along with slightly larger (albeit functionally more capable) form factors such as the MID (mobile Internet device).

And specifically, to that last point, how does this all relate to Nvidia? In spite of the compelling ‘hook’ of the application-diverse x86 instruction set, Intel’s got several substantial barriers to entry in smartphones; a notable power consumption deficit versus ARM, for one thing, which Moore’s Law trends should help mitigate. Another disadvantage is less easily resolved, at least from a technical standpoint; the current sole-sourced status of Atom-based products. Intel offers formidable manufacturing prowess, mind you, thereby addressing supply concerns. But potential customers always prefer multi-vendor sourcing as a means of cultivating competition and thereby securing both feature-optimal products and budget-optimal pricing.

That’s where Nvidia potentially comes in. As I mentioned earlier today, Nvidia’s currently-ARM-only Tegra program hasn’t yet gained traction in the cellular phone market, in spite of several generations of devices and several years’ worth of private and public advocacy to date (not counting the product line’s MediaQ precursors). And, clearly, Intel’s not yet established a beachhead in cellular phones, either. Granted, the two companies are by no means close chums, in spite of their shared aspiration to advance the PC platform (both in general and in competition with the merged ATI/AMD alternative). But in the spirit of ‘the enemy of my enemy (i.e. ARM) is my friend (i.e. Nvidia)’, might Intel end up granting Nvidia an Atom core license, both to secure a second source for its own chips and to keep the FTC off its back?

Jon Stokes also thinks such a scenario is possible. What think you, readers?