That's why AMD doesnt targets 7-8" Android Tablets. But Noooo, Intel needs 40-60millions tablets sales and the only way to do it is contra revenue.
And with Broxton/SoFIA, the BoM problem goes away and all of a sudden these chips you mock Intel for selling "for a loss" suddenly become profitable sockets and a base from which to grow.
When Intel is shipping, oh, 80 million or so tablet chips in 2015, contra-revenue-free, the folks who kept mocking Intel for supporting Bay Trail will finally realize that it was actually quite a smart move.
80 million tablet chips sold at 50% margins at $15/piece is $1.2 billion in sales and $600 million in positive gross margin. That's a winning business strategy crafted by a company with a clear long-term view of what it needs to do to succeed.
Pimping an 11.6" bulky reference tablet with top-bin silicon and highlighting how fast the chip is will do wonders for winning over the "anybody-but-Intel" folks and the "AMD 4 ever" guys, but unless this "superiority" can translate into design wins and ultimately profit, it's worthless.
Let's see what designs Mullins will win and let's see how real shipping devices perform