I recently read about Android’s display model. Apparently its way out of date, always compositing the screen image in software on the CPU. So no matter how good your GPU is, you’re still stuck draining battery and getting lower performance than you could have. In comparison, iOS composites your screen on the GPU, and I bet WP7 does it too.
Defenders of this method say it’s needed because of the low-end Android devices on the market that can’t composite on the GPU. However, the technology to fall-back on CPU when sufficient GPU isn’t available has been around for years. Then the argument is made that having a fallback makes things more difficult for the developer. However, auto-fallback has also existed for years, where the developer writes code to draw something and the OS will decide whether it can be done on the CPU or GPU, and take the appropriate route.
So now you have this awesome chipset, but hindered by the OS.
See this :
Honeycomb fixes that redraw issue on the ARM A15 core, and the NotionInk Adam screen will provide even further power consumption savings for text based usage.
Incredibly I learned about all of this on FR, thanks to Ernest, ShadowAce and the Tech/IT ping lists...