Digging a little deeper into the CPU we actually have a Dual-Core ARM Cortex A9 which runs at 1GHz per core and has 1MB of L2 Cache. This is combined with up to 1GB of LPDDR2-600 and our GeForce has 8 CUDA processors with support for OpenGL ES 2.0, OpenVG 1.1 and EGL 1.4. Tegra 2 can also decode/encode various 1080p formats including H.264, MPEG4, H.263 and VP8.
That seems like a good feature set already but NVIDIA haven't stopped there. Also bundled on the 40nm SOC is a dedicated audio processor (Lossless supported) as well as an image signal processor capable of supporting cameras up to 12MP (with a secondary 5MP) and image stabilisation on video and stills is supported.
Capping this all off is our display controller which is capable of powering two displays, one of which can be HDMI 1.3 with 5.1 audio out.
So what does this mean in the real world?
Well, using the Android browser as an example we have an application which supports multiple threads. With two CPU cores and GPU which supports Flash acceleration in Tegra 2 this means our browser can assign processes independently to the various parts of the chip, enhancing performance and the overall user experience. For example the feel of our browsing is much better as the GPU processing Flash leaves the CPU cores free to handle rendering of the page and other tasks in the background, for example music playback or email synchronisation.
Gamers also get a nice boost in mobile performance too as the GeForce GPU which has its roots in the desktop PC allows more geometry, higher resolution textures and complex shaders when compared to existing mobile solutions. Artificial Intelligence can also be enhanced and we are starting to see games hit the market with Tegra specific content, such as improved character models, more levels or PhysX content which is not possible on older devices.