Intel launches R&D initiative to invent PDA
15 years after Apple's MessagePad, we're still not there yet
IDF Intel today disclosed a raft of technologies that will drive a research initiative it hopes will put the Assistant back into the Personal Digital Assistant, one of the most inappropriately applied acronyms in the history of personal computing.
Intel's plan is called 'Carry Small, Live Large', and while the slogan's new, the concept is decades old, stretching back to erstwhile Apple CEO John Sculley's August 1993 launch of Newton, the operating system and platform on which the first ever PDA, the MessagePad, was based.
Sculley defined a PDA as a handheld gadget smart enough to operate almost autonomously, intelligently linking disparate sources of information for the user's benefit. There was no internet then, at least not for public use, so the Apple CEO was thinking about personal information management. But the notion still applies, and is even more powerful a concept now we have both the internet and the ability to connect to it wirelessly.
While the MessagePad turned out not to be quite as smart as Sculley claimed it would be, subsequent handhelds were arguably even less intelligent, but they all managed to have the term 'PDA' attached to them. They still do, even though the traditional PDA device has now been all but killed off by the smartphone.
None of them have aver delivered anywhere near the functionality Sculley described, but give Intel's researchers five to ten more years and his vision might just come to be realised.
It goes without saying that the Carry Small, Live Large initiative is predicated on shrinking silicon chippery to cram more components into a shrinking die space, allowing handhelds to get smarter and become more connected.
But that's just the customary trend. CSLL assumes all this extra performance can be used to make PDAs more than mere data carriers and readers. So the built-in GPS and Google Maps app know exactly where you are and can feed that into your other apps. The gadget's multi-mode radio is capable of hopping from Wi-Fi to 3G or WiMax (maybe) right at the moment you leave the building, so you're connectivity's not interrupted, ensuring the route the device has planned for you - it knows you have a meeting in a hour - can be changed as soon as it learns there's a big traffic foul-up on the way.
Your PDA's already checked the weather report and told you to bring a brolly.
In other words, the machine's pro-actively assisting you rather than simply presenting information when you ask for it.
This requires fresh smarts - and probably a new operating system too - but it's on its way. Right now it just requires greater performance to allow all the intelligence-simulating data monitoring to be carried out without gobbling up the battery charge. Ditto the real-time video recognition techniques it showed today - right now, these require way more processing cycles than handhelds currently offer, or will for some time.
The 'one-radio, multiple wireless technologies' strategy appears closer to fruition. Intel has built an analogue-to-digital convertor that's capable of sampling the wireless environment around it - it has an integrated spectrum analyser - allowing it to adapt its resolution to meet the needs of the moment rather than assume a worst-case scenario.
Prototyped: Intel's 'digital pre-distortion' power amplifier
Intel's first smart ADC has a 12-bit quantisation and is designed to work with both 802.11n Wi-Fi and WiMax. So too is the 65nm power amplifier it talked about, and we wonder if both will find a home in 'Echo Peak', the Wi-Fi and WiMax radio module Intel will release in May as part of its next major Centrino revamp.
So too may techniques like 'digital pre-distortion' - modifying a signal in the digital domain so that when it's converted to analogue for transmission it's in a form best suited to the wireless environment of the there and then.
This is stuff Intel can do. But while it can demo some other elements of the Carry Small, Live Large programme, it won't be able to implement them on its own. A case in point: having a handheld detect it's in the presence of an HD TV and automatically 'borrow' the screen as its own display mechanism.
Intel showed a wireless chip capable of sending data over the 60GHz band at a rate of around 5Gb/s, but at what point in the image rendering pipeline the data leaves the handheld to be streamed to the TV remains to be determined. Ideally, it needs better rendering technology in the telly, but that it can't do without the help of TV manufacturers, and suddenly its reached the limit of its research - from this point on, CSLL becomes an industry initiative, and a discussion about setting standards.
That gives Intel's researchers more time to work on miniaturisation and power conservation, and to deliver the silicon from which real PDAs can be made. If they're right about the timeline, we could see much of this coming to market by 2013 - just in time for the 20th anniversary of the event at which the term PDA was coined...