OLPC Comes From Venus, Intel Comes From Mars

Intel and OLPC have gone their separate ways, and it has not been a beautiful thing. You can see OLPC’s statement here, you can see an interview with the head of OLPC here

In return, Intel said it left because of “philosophical differences.” apparently evidenced by OLPC’s insistence that Intel stop competing against OLPC’s XO with its own designs.

It appears both are correct. Two quotes from OLPC sources get to the heart of the matter:

“[W]e view the children as a mission; Intel views them as a market.”

“We’re like the World Food Program and they’re McDonald’s. They can’t compete. . . . Paul Otellini himself would say that one size doesn’t fit all. And we agree with that, of course. That wasn’t the issue. They cannot compete with OLPC and be a partner.”

It may not be PC to say this, but in the long-run, Intel is right and OLPC is wrong, and the world, including the poor kids, will end up be better off with Intel winning

We’ll explain.

The OLPC XO is an idea whose time has come . . . and gone. It is a victim of its own success.

What do I mean by that?

OLPC succeeded before even building a single unit by drawing attention to the fact that what the less-than-rich world needs is a simple, cheap, portable computer.

They’ve served as a tripwire to get the biggest CPU company in the world to shift attention and make providing that a priority with its Silverthorne project.

Unfortunately, having done that, OLPC now refuses to step aside, and that’s the problem.

The sad reality is both the OLPC XO and Intel Classmate are essentially cobbled-together compromises between too-old and too-new technologies. They are going to look terrible a few years from now compared to what will be available then.

The XO is essentially an ancient 450MHz processor with insufficient storage space. The Classmate is a 900MHz processor which chews up too much power and also has insuffient storage capacity.

In comparison, by 2009/2010, Silverthorne products will deliver the equivalent of a 1.6-1.7GHz Pentium M, and flash costs will be low enough to provide a reasonable amount of storage at a low price. It will be the functional equivalent of the lowest-end $400-450 notebooks you see today, and will probably cost a bit less than $200 in volume.

Such a machine not only looks like a much better deal not only for any poor kids, but for their poor parents, and lots and lots of non-poor folk in non-poor countries.

Not surprisingly, Intel sees this as PC 2.0, and sees this as the key to dominating the computing world for the next few decades. No doubt they viewed joining OLPC as a means to market this, and anything they come out with in the meantime, with the XO being relegated to bottom-of-the-barrel sales and eventual phaseout.

And what do the XO people say? They say, “Don’t do it! Stop building such machines on your own!! If you must, give the technology to us so we can get all the credit!!”

OLPC comes from Venus, Intel comes from Mars.

Really, what does OLPC really want to be, a charity or a computer company? This is like the American Red Cross finding out that WiMax would be great for emergencies, then demanding that they should be the only ones making WiMax devices.

Sorry, not going to happen.

You may say, “But the XO can get better.” Sure, but who is going to do it? AMD? Let’s get real. They’re cutting back on R&D for their profitable lines of business, do you think they’re in any shape to spend any serious money on this anytime soon?

The sad reality is Silverthornes are going to wipe the floor with the XO. While they may end up costing more (and maybe not, as the Asus EEE shows, there’s a much bigger market for these devices than just poor kids), they’re likely to be a much better long-term value. That will make anybody buying XOs in 2008 look pretty stupid in a few years, which will lessen XO sales considerably the next couple years.

And there is nothing, nothing, OLPC can do about that. Tossing Intel out of their club is just urinating in the wind.

Unless OLPC wakes up from its fantasies of becoming some sort of anti-Dell, an Apple for the poor, Intel will crush it in a few years, and on balance, that will be a good thing.

Unfortunately, something will be lost as a result. What OLPC needs to do is to realize that they’ve unconsciously, unintentionally succeeded in ensuring that cheap computers will get to the poor soon (if not quite as soon as they would have liked). They no longer need to build them, what they now need to do is make sure they’re used, and used effectively for education. OK, maybe they can help governments get lower prices for the boxes, too.

It’s a less glamorous but more important mission. Carpetbombing poor countries with computers isn’t the mission, it’s only the start. In the end, the hardware really doesn’t matter, it’s how or even if it gets used.

The poor world is littered with innumerable aid projects started with the best of intentions, but later abandoned as the donors moved on to the next fad, and the donees didn’t pay to keep it up.

Good intentions aren’t good enough.

Whether OLPC or Intel ends up making the hardware, there is a great danger of that happening. If OLPC is really, truly interested in helping the poor with computers, rather than making themselves feel and look good while empire-building, they ought to stop competing against technological powerhouses and instead change the focus to making sure the boxes are used to get the job done.

They ought to forget about the nuts-and-bolts and let Intel or AMD or Via or anybody else build the boxes, and dedicate themselves to making them work.


Be the first to comment

Leave a Reply