ARM's hard macro version of its Cortex A9, called Osprey, has been received by many as the processor that will challenge Atom for the Netbook/Smartbook/MID/UltraMobileComputing (whatever you like to call it) market.
Osprey, built to run on TSMC's 40nm high performance process, won't, reckons ARM, be in products for a year. That means September 2010. By then, of course, Atom will be on 32nm.
Osprey runs at a maximum speed of 2GHz. Intel is at 1.6GHz today. It's safe to assume that, by this time next year, Atom will be faster.
The big deal, of course, is power consumption. ARM says its chips allow a Netbook to run all day. That's also the claim for the Atom-powered Nokia Booklet. Both camps are bringing power down fast. ARM is well ahead, but a determined Intel with a newly empowered Dadi Perlmutter, its low-power evangelist, may not lag forever.
Intel says it will put its GPU on-chip which will reduce power and increase performance. Will ARM add its Mali GPU to Cortex? "It seems reasonable to aspire to that", says ARM's John Goodacre. A less belligerent stance than Intel's.
Then there's the issue of the surfing experience. Intel says that x86 gives a better experience, but the ARM-based iPhone 3GS seems to be just as good to me. While, of course, Adobe has completed its
Then there's the Linux vs Windows thing. Is this the show-stopper for ARM-based Netbooks?
People prefer Windows. But forever? Some see Google's Chrome OS, delivering Google application software, as the answer. But that's just another aspiration. Meanwhile one can't see Microsoft porting Windows 7,
The heady prize for the world's computer users is that ARM-based Netbook processors are said to be coming from TI, Freescale, Nvidia, Qualcomm, Samsung, and Marvell. What a treat after the 20 year-old bipolar x86 supply situation.
But the optimism of Spring seems to have turned chilly. It's all taking a bit longer than expected, the questions don't all seem to be getting answered and the certainties don't seem to be all that certain any more.
Success of Osprey over Vulture will most likely depend on the market acceptance of the business models of the two players.
Will the market(s) accept another Intel fat margin (not now, but later) dominance in these new mobile sectors? I doubt it. For now Intel can buy the business, but can they sustain that model?
Add in the cloud and its impact on the nature of applications and we'll see both players scrambling for several years yet to keep pace with the market re-organizations that will, indeed are, occurring.
My prediction? There is space for both players here for some time yet, and I doubt either will have a significant dominant position over the other in 3 years time - because that competition will suit the market.
Geoff, I'm sure you're right but it looks as though Intel will fight like a tiger to hang onto its control of this market.
I just have to say how great I think your blog is. You have some of the most level-headed commentary in the tech blog scene, and you talk on such a variety of subjects that I don't always see elsewhere. It's sad to say, but a good tech-blog is a rarity in this technological day-and-age!
I think the computing market is at the point where people - although used to seeing Windows on traditional computing devices - are more aware than ever of alternatives. Most of the major high-end phones sport alternatives to Windows, and I don't see people incapacitated by the user experience. Although they may think that computing means Windows by default, the advantages - and similarity - of open source alternatives could soon become apparent to the masses, provided that retailers can also come to understand those advantages as well.
As computing undergoes its current upheaval, the nature of the devices and the way that they operate are changing, and people could become more open - in my opinion - to seeing and seeking new interfaces. So long as an operating system is not too drastically different to what people know, then they can muddle their way through. The interoperability inherited from the mobile world will hep make this a less painful experience than it has been in the past.
I would agree with anyone who says that the open model isn't perfect yet, but the market is getting very close to a viable alternative to some of the traditional ways of computing, and it is perfectly possible that an open model will prevail; this might leave closed, proprietary systems (such as Windows) struggling to outdo the open source community, who may ultimately overtake it once a developer "tipping point" is reached.
So my optimism hasn't chilled one iota, since I can see already that the old guard of the computing industry is rattled, and they no longer have the party to themselves.
First Alex, thanks for the compliment, much appreciated. Second, let's hope so. The thought of Intel and Microsoft dominating this thing for ever and ever is just too depressing.