Originally Posted by TymonFlak:
I'll bet you are already seeing the 3D hardware in use. Apps all render to their own space, when you 'push the view back' to card view. apps continue to render to their 'full size' space. The CPU is not drawing all those pixels at a lower resolution while in card view. It's using the texture filtering of the integrated graphics to smoothly 'shrink' the image.
All the transparent effects are also using the accelerated blending of the graphics as well.
I also would not be suprised if all the html5 <canvas>(path, rect and image) methods are HW accelerated.
The powerVR onboard is for polygons and depth buffering. it doesn't do transforms, but that's what the DSP is for.
The pre is effectively a 'multi-core' device. The CPU can be doing stuff, like slurping network data, while the DSP decodes video, and the powerVR filters or rotates the image... all at the same time.
Sounds good to me!