And you are being unfair with me. What I've been saying to you that Kabini didn't hold the kind of marketing advantage Brazos had at the time, so it would have a *much* harder time on the market, as it wouldn't have a niche where it ruled, but just a very small gap between intel cost (Atom) and performance (core) parts. This prediction held true: Kabini sales didn't take off, we got a small uplift in Q2 and they again went back to normal and started to bleed market share again. And now the ultimate joke: AMD is trying to build a business case for Kabini on the desktop.
It was your prediction, that AMD would be able to "flood the market" with Kabini parts that didn't held true. Despite selling at below 40% margins, AMD can't make sales take off, while Intel will be able to keep around 60% margins with 22nm Silvermont. This, as I told you, happened because costs on 22nm Atom are better than Kabini's, and the chip is also more efficient.
But since you are talking about my predictions here, mark these words: AMD will bleed market share on the bottom market from Q114 until the foreseeable future, because Bay Trail, Haswell will be able to compete on costs and the 14nm chips will eat AMD line up for breakfast.
I said in 2011 on SA that AMD would have to eventually quit their x86 business because they wouldn't be able to recover from the Bulldozer disaster, and while they didn't quit, their x86 business is now a former shadow of itself, out of the growing axis of the "new AMD". Who was closer to the target, the guys expecting miracles from Brazos, Steamroller (the chip that was what Bulldozer was supposed to be lol) or me?
Ed: Krumme, what happens if Microsoft decides to backport the DirectX 11.x API from the XBO to the PC, giving straight compatibility with AMD hardware and access to a lightweight API for Intel and Nvidia hardware? R.I.P. business case for Mantle.