Any of you that's believing that we'll see Atoms move to 22nm the same time as their PC chips(early 2012) is in their own dream world. Think about it. Late-2011 is the FIRST 32nm Atom products. Assuming the consumers can actually buy it at that time.
1. How many times have Intel skipped a process technology generation? You do not skip a process technology generation, unless its hopelessly behind like the Itaniums and have no choice. 2. Even if they didn't skip it, and even considering 32nm will only last 1 year, that still means late-2012 for 22nm, which is perfectly fine, but very optimistic. 3. The CE chips were the first non-Netbook chips to be available. Somehow we're betting on an unreleased chip to be on the process tech that's not even released yet. 4. When Intel says 32nm was coming in late 2009, that means you could buy the products based on it. It doesn't work on smartphones, because the "infrastructure" or whatever is required to bring the product at the same time as the hardware takes time.
Let's see how aggressive Intel needs to be just to be competitive.
Nvidia's Tegra roadmap showed the following: Tegra 2: Fall 2010 for Tablet, Early 2011 for Smartphone, which was correct Tegra 2+: Spring 2011 for Tablet, so assuming 6 months difference smartphones will hit around Jul-Sep. Tegra 3: Fall 2011 for Tablet, looking at Tegra 2, it'll be early 2012 for smartphones
Single thread -Lincroft 1.5GHz: 500 -Cortex A9 1.0GHz: 245
Intel will need a dual core version of Medfield for smartphones in Fall 2011 to have a headroom to be competitive with the quad core Tegra 3's coming in a few months after that. You know, while achieving the form factor, power consumption, graphics, software, everything.
Market Data 
Markets 
AI
