From what I can tell, the whole thing is basically a PR stunt. Intel are less interested in actually displacing ARM than they are in generating lots of publicity about how they're part of that hip new IoT thing. They've funded an entire TV show called America's Greatest Makers to try and associate the Intel Curie with wearables and the maker movement, done a whole bunch of high-profile tie ins with the X Games, Lady Gaga, Arduino, BMW, etc but they've no interest in actually making it available unless it'll give them publicity.

It seems to be impossible to find official power consumption figures for it anywhere either, and the documentation is lacking in general.

The Curie is a nice little SOC and it will be available to purchase. If you read the press release associated to the recent internal reorg at Intel they mention IoT as one of their focus markets.

Intel isn't "Papa's Fabless Semi Company", and they've got better things to be doing with their expensive fabs than running chips through them to sit on shelves for the rest of time, hoping that someone comes along and buys them. I completely understand why Intel's not interested in wasting a ton of money doing a big production run of these chips, and then having to worry about things like the product's lifecycle without a large enough financial incentive to back it.

That isn't to say "they won't sell their solder on chips to distributors," because they absolutely do: Go to Mouser, select Intel, and look - Quarks and Atoms galore, even high end networking chips, all available for purchase.

One of the sad things I've learned first hand about the tech business is that even profitable small products are just not interesting to billion dollar companies. You need large enough revenues to make a splash as well, and these tiny MCUs selling at quantities of 10K here or 20K there just isn't enough to make Intel want to work for it.

Furthermore, if Intel wants to compete with these SoC vendors, they need to start acting like them - the Curie being a good example of that. They built a custom chip to do what their customers asked of them, and now they're selling it directly to those customers. Everything they're doing is fitting the strategy they've laid out.

> One of the sad things I've learned first hand about the tech business is that even profitable small products are just not interesting to billion dollar companies. You need large enough revenues to make a splash as well, and these tiny MCUs selling at quantities of 10K here or 20K there just isn't enough to make Intel want to work for it.

This is one of the core thesis of the Innovator's Dilemma: the multi-billion dollar companies aren't interested in competing to serve your niche application.