Intel launches new Atom family for mobile devices

Hello Alistair,

in my opinion Intel wont stand a chance to gain market shares worth mentioning within the next years.

Especially the mentioned agreement with Nokia doubles my point of view. The now presented and available devices are just running under Moblin and not, as expected, under MeeGo. And most of the acception on the side of customers will come with an open app store with a sufficient number of applications.... without thinking about different versions of OSs. On the Computex Intel talked alot about new devices planned by all the "big players" on market... but most of these, like eg. LG with their GW990, stepped back.

Sure we all might be excited by the possibilities of the hardware But on a smartphone or tablets with limited access to power... all these "extras" are consuming too much... first of all the 3D graphics. Even if the sleepmodes SOi1 and SOi3 of the platform are very remarkable for a x86-system... the whole platform with all needed ICs still consumes about 21mW... alot for the deepest sleepmode.

I just dont think that the fixation on the x86 architecture will grant success in this market... as for the competition alot of mature concepts, operating systems and complete solutions are already available... even more as Microsoft also retreated as Windows wont run on these devices at all.

Yes its an impressive piece of engineering But placed in the wrong market segment (my opinion).

Hi Christian,

Thanks for the reply.

The chip does seem a bit power hungry for mobile applications, but I believe when the geometry is shrunk from 45nm to 32nm the power use will drop quite a bit. As always it will be a balance between user demands for more performance and battery life. I think I'd accept the ability to play a hardware intensive game when I am away from home without a laptop. For instance, when I am on a train back to Scotland which lasts four hours, I'd gladly charge the phone up when I got home if I could have four hours of decent gaming on the train. As a comparrison, I can get only get around 90 minutes of playing Civilization on my laptop at the moment.

Is Meego operational yet? I've got to be honest and say this area isn't my strong point. So I did have a browse online and found this blog http://meego.com/community/blogs/valhalla/2010/towards-day-one. The author says day one would be when they open the Meeg repository at the end of last month. I'd imagine either they had trouble running Quake on Meego as it it is for the demo, or they were saving Meego for the official launch

What market do you think Intel should have aimed for? All that power seems overkill for run of the mill applications like dashboards etc, and I'm not sure it would be powerful enough for really high number crunching applications like mediacal scanners. Maybe netbooks? I think those will be the first market ready applications

Hi Alistair,

I agree that power consumption will go down quiet a bit when the shrinking process is done. But even then, in my opinion, there will still be a big gap between ARM and this platform. Not only because of the sleepingmodes but most of all because of the consumption under heavy and normal load. I was quiet surprised when the former plans to implement 2 independent graphiccores have been abdoned... so for now you dont have the possibility to disable the 3D-part completely... and this way you ll need way more power than would be necessary.

Yes, I understand your point of view for being able to play... but do you really think that you ll be able to enjoy (!) 3D-effects and acceleration on your smartphone ? By the dimensions of a normal smartphone-display you wont even recognize most of them and after a few hours your eyes will regret your decision Thats the segment of more specialized gameconsoles... and alot of them already lost in market because of small displays. Thats my point: on smartphones these possibilities wont bring big advantages as these devices primarly have been designed for other applications and communication. Cores with the now available computing power are more than capable to handle the required calculations themself to bring a 3D-effect on these small displays (even if the resolution might be quiet high... this wont be used in games anyways). With mobile devices like Pads, Netbooks or Tablets I agree... but Intel highlights especially smartphones...

As far as I know Meego is not running properly (supporting all features) on this platform yet... same with the only second possible OS... Android. And as this OS should have been a big plus of the new platform, the time to present this was just not right in my opinion. Its like you would present a hybrid vehicle with just the gasoline engine working by then... If they would have waited with the lauch, a whole package would have had more of a "bomb"-effect.

The biggest advantage for bringing the X86-architecture to mobile devices would have been the easy portability of already available applications to your new devices... and to be able to continue working within your familar programs. But now there will be just another "closed" area... with new developments being necessary. Yes the systems are based on Linux... but you still cant port applications directly. Not to mention Windows... as this will never run on these devices because of the lack of a BIOS or EFI... at least as alternate possibility this would have granted a big ressource of already running applications right from the start.... the main condition if you want to stay a chance against an already available and common platform. And again for games... how many are already available (with 3D) and how many gamestudios will adapt them for the new platform ?

The market for these devices is widespread... and in my opinion smartphones just should be at the end of the list right now. Netbooks and Tablets are ones of the first that I think of. But also graphical interpretations of eg measurements in complex systems, automotive applications, and general automation comes to mind. My favorite would be (a little biased ;o)) homeautomation... wireless devices with streaming capability and access to the complete control systems. One interesting example (not my company, so no ad ;o)): www.eyecontec.com . And as this market is still at the beginning and has great potential I just think this would have been a better point to start of.

This all might sound as I am all negative about this platform... I am not. Just a bit dissappointed in some areas and I just dont agree with the orientation of Intel in some ways.

I really can't see anything technical to disagree with in your post, but I will say that you are looking at the present generation of smart phones. I don't think Intel are aiming at that market. With current design cycles it will probably be at least a year before we see a mobile device featuring this generation of chips, and 18 months till the smaller geometry version sees the light of day in a commercial product.

If I look back 18 months it would be probably be difficult to imagine phones with the features of the HTC Desire/incredible for example. What new features will top of the range phones offer in 18 months?

The Atom may not compete powerwise with current ARM processors, but ARM can't compete with the features of the new Atom. There will be a newer ARM release that will probably look a lot like the Atom, which is intended to compete on the same designs. I reckon it will give designers a real choice, is power or features the main concern? Will newer batteries/fuel cells offer the power to keep Atom based designs to a level that will keep consumers happy? How can a mobile provider provide a full range of phones that will consist of top end design with lots of features and lower end designs that offer basic functions and long battery life, while still staying code compatible?

I think it is a really interesting market, and am looking forward to the next couple of years.

You are correct also about mobile consoles, and I have a PSP, but I would rather have a phone with similar capabilities. Every time I travel, I end up taking a second bag just for electronics and entertainment; PSP, phone, iPod Touch, Laptop, books, and all the related chargers, headphones and other accessories. One device that had a decent browser, could make phone calls, had good media storage and high quality repoduction, a games capability at least equal to my 1986 Commodore Amiga, Outlook synced, and a 50% larger screen than current smartphones would empty my savings account .

Not being able to disable 3D and no independant graphics cores does seem a bit fail. Hopefully the later iterations of the chip bring those back