I’m a big fan of the x86 platform when it comes to mobile Internet. To me it’s simple, if the Internet is written on x86 platforms, how can we expect it to look the same and keep up with the latest techniques when its used with ported software?
However, this is the mobile Internet we’re talking about here. A new and different world where customers are just starting to move beyond the idea of SMS’s. Right now, today, even a basic, cut down, tailored and slow experience is something new and exciting and if its mobile and consumer oriented it better be small and stylish otherwise it doesn’t even get past the first hurdle.
That’s why today, the only option for consumer handheld Internet is ARM based. If you’ve read any of my ARM-bashing articles on UMPCPortal you might find that a bit surprising but there are different requirements for different markets. The UMPC market is very much about productivity and full desktop capability. The consumer mobile Internet space is more about getting Facebook and YouTube working!
A Low-power, low-cost, small sized and a reasonably tailored Internet experience is going to be fine for the next 1-2 years for most mobile Internet consumers and that means that the smartphone/featurephone platform is going to be the seed point. Later on, when dedicated devices running on Moorestown (I don’t see Menlow being small or efficient enough.) or on variants of VIA’s mobile-ITX platform mature then there might be an opportunity to take advanced mobile Internet consumers to the next level but we’re talking about 2009 in that case and for the time being, ARM is in the best position. And they could stay there too. If the platform processing power increases and the software development activities speed up then there’s no reason why ARM can’t stay at the spear-point of the consumer mobile Internet boom. Last week we saw two announcements from ARM and today there’s another very interesting announcement from Mozilla (via a twitter from the jkOnTheRun team) that I will analyse in a separate post. They all point towards a huge effort on the software and processing power element and that’s exactly the right sort of focus.
Maybe, in 2009, there’ll be another market split where we see low-end mobile Internet consumers staying on tailored platforms and client software. Possibly where they have been tied in to contracts that bind them to certain platforms (this could be a good strategy for ARM!) and then we’ll see the people that do serious Internet work (and play) moving to dedicated devices on an x86 platforms. Personally I think there’s a huge cake there and both teams will win in the medium term but as we move forward past 2009 I can’t help thinking that the reduced development costs/duration of x86 solutions, along with the eventual insignificance of the power drain differences are going to give the x86 camp the edge that’s needed for long term future success.
October 11th, 2007 at 6:03 pm
what i suspect is taking time on arm based devices is that they insist in reinventing the UI rather then just slap windows on the device and leave it at that (complete with the issues of how useless the windows ui is on a small screen).
hell, it seems that rather then selecting a existing distro, they insist on creating their own distro, over and over…
i just damn well hope that ARM plan on including hildon support in their distro, as that will make porting stuff from maemo (nokia) as simple as a repack at best.
hell, it may well be that if done right one can use maemo packages on whatever ARM comes out with.
and didnt intel list hildon support as one of the requirements for moblin compatible distros?