Intel began porting Android
Intel’s president and chief executive Paul Otellini says his company is hard at work porting Google’s tablet-specific Android 3.0, aka Honeycomb, to the x86 architecture.
“We’ve received the Android code – the Honeycomb version of Android source code – from Google, and we’re actively doing the port on that,” Otellini told reporters and analysts during a conference call on Tuesday announcing Intel’s first-quarter 2011 financial results.
Otellini’s admission put the stamp of reality on recent rumors that Intel is working with “first-tier notebook vendors” to build tablets based on Intel chips. It may also shed more light on why Google has yet to open source the tablet-centric version of its “open source” Android operating system.
After Otellini added that “[we] expect to be able to ramp those [Honeycomb-based] machines over the course of this year for a number of customers,” he was asked how well the move into smartphones was going. To this question, Otellini was specific: “I would be very disappointed if we didn’t see Intel-based phones for sale 12 months from now,” he said
Two years ago, Otellini told a gathering of investors that Intel would be going after the smartphone market, saying that handsets, along with other mobile and consumer electronics devices, were “what we’re aiming at. This is where we think the growth opportunity is for us.”
Two years after that pronouncement, Intel hasn’t yet cracked the smartphone market – nor, for that matter, the tablet market that Apple created essentially singlehandedly with its “magical and revolutionary” iPad, which shipped one year after Otellini told investors about his company’s “growth opportunity.”
Not that Intel has been idle in the mobile marketplace, Otellini reminded his audience. “We…launched Oak Trail last week, which is a platform designed specifically for tablets. We are seeing very good design momentum with Oak Trail across multiple operating systems. Over the course of this year, Intel will have tablet platforms that run Windows, Android, and MeeGo.”
But tablets aren’t smartphones, a market in which Intel remains stalled while ARM rules the roost, and a market in which Intel is banking on its upcoming 32 nanometer Medfield processor. “We remain committed to success in the smartphone segment,” Otellini promised Tuesday’s listeners, “and we’re actively working with a large number of handset manufacturers and carriers around the world on Medfield-based designs.”
Intel has been smacked around a bit recently in the smartphone market, Otellini admitted. Referring to Nokia’s decision to go with Windows Phone, Otellini said: “In terms of phones, obviously we lost Nokia, which took a lot of the wind out of the sails for phones this year.”
When life gives you lemons, the old saw says you should make lemonade – and Otellini says that’s what happened with his Nokia-targeted team. “We’ve redirected those resources onto a number of other major accounts, focusing on carriers who want their own devices, and also on handset manufacturers,” he said. “They’re all based on Medfield, which I think is still the first 32 nanometer phone apps processor in the industry.”
When asked what would drive interest in Medfield among Intel’s customers – MeeGo being open source and customizeable, or Medfield’s competitiveness with the ARM architecture – Otellini answered: “It’s both. Early customer activity is on Android and on MeeGo, and we’ve got ports on both those operating systems.”
He also seized the opportunity to bang the drum for Medfield, calling it “very good in terms of performance, particularly in the area of media – high-def video replay, that kind of thing – on phones. And the power is right smack dab in the envelope that you want it to be.”
He also went straight after ARM when dodging a question about how the x86 architecture would be of advantage in a smartphone. “In terms of x86 versus ARM, y’know, it’s not just about the core, as much as we would like it to be and as much as I guess the ARM guys would like it to be. It’s about the core, the overall capability of the system-on-chip, the things you put around it – the graphics, the comm subsystems, the media-processing subsystems – and the overall power envelope relative to the performance that you can deliver of the SoC.”
He also downplayed that fact that as of this moment in time, no true smartphone OS runs on x86. “I’d also point out,” he said, “that all of the major operating systems in phones – in smartphones – are written at a high level, such as they’re cross-platform and portable. And so it is easier for people to move from ARM to Intel, or ARM to ARM, than it has been in the past in the Windows world.”
That may be true, but there has been little incentive for anyone to undertake those porting efforts, seeing as how Intel hasn’t yet come up with a platform that will fit within a smartphone power envelope. That’s been ARM territory, and it will be until Medfield ships. And maybe – likely? – after.
But now Otellini and his company are working feverishly to ensure that 2011 will be the year when Intel Inside™ might be a sticker you’ll find on a tablet running Andrpod 3.0, aka Honeycomb.
For an Intel Inside™ smartphone, however, you’ll need to wait until next year, at least – three years after Otellini’s 2009 pep talk to investors. via TheRegister