next »

[eduardomontez]

I wonder if the XO-2 is going to use the ARM processor (en.wikipedia.org/wiki/ARM_architecture).  From what I understand it uses a lot less electricity than even the Intel Atom, and has a far smaller number of transitors. 

Ubuntu has been ported to it, so Fedora presumably could be.  And XP won't run on it at all. It would seem to fit in with Pixelqi's strategy of building all the computing power into the display. So does this seem like a reasonable idea?

[teapot]

I wonder if the XO-2 is going to use the ARM processor (en.wikipedia.org/wiki/ARM_architecture).  From what I understand it uses a lot less electricity than even the Intel Atom, and has a far smaller number of transitors. 

I am surprised that ARM wasn't used in XO -- Sugar should work on it without any problems (traditional "netbook" applications may need x86 compatibility for binary media codecs, however Sugar doesn't need them).

Ubuntu has been ported to it, so Fedora presumably could be.

Porting Linux distributions to different CPU architecture is a nontrivial effort, so unless Red Hat wants to support the whole distribution on ARM, there would have to be a separate port of Fedora subset used in Sugar.

And XP won't run on it at all.

I am sure, Microsoft would insist on running Windows Mobile on it.

It would seem to fit in with Pixelqi's strategy of building all the computing power into the display. So does this seem like a reasonable idea?

I am more concerned about touchscreen as a keyboard replacement -- what works well for phones is not necessarily suitable for typing long essays.

[eduardomontez]

Yeah, the touchscreen bothers me too.

[eduardomontez]

Will Apple make a netbook with the ARM chip?

http://blogs.computerworld.com/apples_tablet_to_be_based_on_arm_cortex_architecture


http://blogs.zdnet.com/Apple/?p=2619

[Gabey8]

Will Apple make a netbook with the ARM chip?

http://blogs.computerworld.com/apples_tablet_to_be_based_on_arm_cortex_architecture


http://blogs.zdnet.com/Apple/?p=2619

I'm reading through your first article now. I see a huge motivation for OLPC to at least look into ARM as a possibility, if only for the extended battery life it could provide. Here's a quote from it regarding the use of ARM chips and their effect on battery life. Where have we heard THIS phrase before? (Emphasis added by me.)

While pretty close in speed, the ARM Cortex A8 absolutely kills the Intel Atom in power efficiency.  We are talking Apples and Oranges here.  Not double, but an order of magnitude better for ARM.  Some ARM chips routinely use 10-20 times less power than Intel for similar operations.  Battery usage with ARM chips in prospective netbooks could be measured in days, not hours – much like smartphones.

[ChristophD]

When it comes to the important metrics (price, performance, power-consumption) the gap between ARM and x86 is somewhat closing these days. However I agree that ARM will probably have a slight edge here in 2009 so I wouldn't be at all surprised if the XO-2 weren't x86 based anymore. On the other hand the software ecosystem simply isn't there yet at the moment. As mentioned above some significant efforts are underway but there's still a long way to go.

2009 shall be an interesting year indeed...

[eduardomontez]

Freescale announces reference design for $200 ARM netbooks with 8 hour battery life.

http://news.zdnet.com/2424-9595_22-257347.html

How close can Atom come to this?

[eduardomontez]

Oops, I forgot to mention it would run Ubuntu.

More information here:

http://media.freescale.com/phoenix.zhtml?c=196520&p=irol-newsArticle&ID=1240267

[eduardomontez]

A key question here is how the XO-2 can get low enough in power consumption so that it can be powered by a child. It looks like ARM can do the job, but what about Atom? At present it is way too high. However, according to en.wikipedia.org/wiki/Intel_Atom the problem is with the chip set, not the core, and next year it is going to a SOC (system-on-chip) design that presumably will make power consumption much lower. I am wondering if it would be low enough to be child-powered.

[ChristophD]

eduardo, the problem with the Atom platform is still that it's relatively expensive and will probably remain so as Intel doesn't want to lower the average sales price of their products even more...

I think that the device price might be the singlest biggest differentiator between ARM and x86 low-cost when it comes to small and portable devices in 2009+.

[mavrothal]

I see...
Atom is too  expensive, ARM is not x86, VIA C7-M ULV is?... (too power consuming maybe?...).
We'll just hope for AMD to shut down the Geode line to see a change in the XO...

[eduardomontez]

There is another matter here, which is OLPC's plan to run Windows on the XO.

But first let's start at the begining. OLPC wants to reduce the cost of the XO-2 to a hundred dollars or less.  A key part of this is that Pixel Qi is planning to get rid of the motherboard and make the computing circuits part of the display.

Now ARM fits right in with that plan, since it is simple and the ARM business model is one of licensing the design to chip manufacturers. On the other hand, I don't know if Intel would even license Atom, and from what I understand it is more complex and expensive to produce, and furthermore there are concerns about power consumption. So it seems ARM is the way to go.

The problem is that OLPC is porting XP to the XO, at least in part because a lot of developing countries say they prefer it. But XP doesn't run on ARM. And furthermore Microsoft is going to come out with Windows 7, which, as far as I know, is not being ported to it.

I am not sure if I have all the facts right, and in any case there are probably going to be some unexpected developments. But at this point in time it seems like OLPC has to make a decision between going with Atom and blowing the cost goals, or going with ARM and losing a lot of countries as customers.

One more thought: this year we are going to be seeing $200 netbooks runing Linux on ARM.  My guess is they are going to sell like hotcakes in the developing world. If that happens then it would tend to tilt the balance in favor of ARM.

[teapot]

The problem is that OLPC is porting XP to the XO, at least in part because a lot of developing countries say they prefer it. But XP doesn't run on ARM. And furthermore Microsoft is going to come out with Windows 7, which, as far as I know, is not being ported to it.

Not a problem for our Redmond friends! They have an ARM-based product, Windows Mobile. Currently its a** is in the process of being kicked by iPhone and Symbian, and likely things won't get better if Android will become popular, so I am sure, Microsoft will try to present that abomination of an OS as their new platform for XO.

[eduardomontez]

You may well be right. Microsoft is going to have to respond to the $200 ARM Ubuntu netbooks. It seems to me most likely it will port Mobile to them, and in that case, OLPC could produce an ARM X0-2 and have Windows running on it. Or maybe Microsoft will decide to go some other path. One consideration here is it likely doesn't want to undermine the ClassMate by supporting ARM, but who knows?

Part of this is a problem of timing. So much of the hardware and software design depends on the cpu, so it seems to me OLPC is going to have to decide real soon, and before it knows what Microsoft is going to do. But maybe I'm wrong about that.

[joncamfield]

Well, it's not going to run on AMD's Geode anymore:

"Advanced Micro Devices on Monday said it has no replacement planned for the aging Geode low-power chip, creating uncertainty for its use in products like future XO laptops made by One Laptop Per Child.

There won't be a Geode successor and the company has no core microarchitecture planned to replace the chip, AMD executives said. The comments end speculation about the future of Geode, an integrated chip used in netbooks like OLPC's XO laptop, ultramobile PCs and devices like set-top boxes.

"There are no plans for a follow-on product to today's available AMD Geode LX products, but we expect to make this very successful processor available to customers as long as the market demands," said Phil Hughes, an AMD spokesman.

The chip is too old for further development, said Dean McCarron, president for Mercury Research. Chip designs and manufacturing processes have improved since it was first introduced.

--http://www.pcworld.idg.com.au/article/274414/amd_sees_no_geode_chip_replacement_sight

At least, no "new"  geode - so it's stuck with the performance it has or will need to move to a different chipset.  Probably not high on OLPC's to-do list, in any case.