The next version of Google Glass, which is due out sometime in 2015, will be powered by an Intel x86 chip — rather than the Texas Instruments ARM chip that helms the current Glass Explorer Edition hardware. This will mark the first major hardware revision for Glass, which was first shown off at Google I/O 2012 and went on sale in limited quantities in early 2013. Presumably this is either to boost the performance and battery life of Glass considerably — or Intel cut a very generous deal that encouraged Google to move away from ARM for its wearable computing efforts.
Currently, the Explorer Edition of Google Glass (which is available to public as of May 2014) costs around $1,500 — which is a pretty steep price to pay for a device with limited battery life and an ancient TI OMAP 4430 SoC. When the consumer version of Glass finally arrives, priced at around “the average smartphone,” it will ideally need all-day battery life — a problem, when you’re dealing with such a tiny form factor (it’s not ideal to carry a large battery over your ear). It would also be nice if future versions of Glass had built-in cellular connectivity (at the moment they need to be paired over Bluetooth with a nearby smartphone). Again, though, this would be a big drain on battery life.
According to a report from the Wall Street Journal, citing people familiar with the matter, Intel will supply the SoC in the next version of Google Glass. It isn’t clear if this will be the consumer version of Glass, or merely an updated version for developers/early adopters. Sadly, there’s no word on what Intel chip Google has chosen — but given how the current OMAP 4430 SoC is a power-hungry 45nm part, I would’ve thought that many of Intel’s newer 22nm and 14nm parts would potentially fit the bill.
Given the 2015 release window, and Intel’s work to make Android tablets work well on x86, a 14nm Atom chip would probably make the most sense –Cherry Trail perhaps, or maybe Morganfield. The other possibility is that Google might decide to dip into one of Intel’s newer, edgier, ultra-low-power chips — Quark and SoFIA. SoFIA, an x86 SoC with integrated cellular radios and built at TSMC on 28nm to cut costs, could be a particularly interesting prospect — if it has enough grunt to drive Google Glass, anyway (I think SoFIA is more intended for very low-end smartphones and the internet of things).
But why is Google moving to Intel x86 in the first place? Wouldn’t it be a lot easier to just switch out that older TI SoC for a newer Qualcomm part? Moving to x86 isn’t that drastic, but it will still cause more upheaval (developers, tooling, integration with the greater Android-ARM ecosystem) than is really necessary. Google must’ve had a very good reason to switch to Intel — either Intel has some exciting low-power SoCs coming down the pipeline, or it cut Google a very good deal. Obviously, if the next version of Glass is angled more towards consumers, it would be a very nice PR coup if every device was emblazoned with Intel Inside.
Source:Extremetech