[JesseK]

Allow teams to bring a [computer] (ITX form-factor with a NVIDIA GTX 8xxx card in it) to the field. The CPU will perform all external interfacing and memory management while the GPU will perform all FOP's necessary for true high-resolution pattern recognition algorithms via the vision system. The libraries for GPGPU are now mature enough for us to be able to do this. Of course, the overall computer would account for ~$1k of the cost of the robot so there may or may not be an advantage for a team to do this.

Maybe I have misinterpreted what I remember from the Atlanta presentation regarding the intended external interfacing on the driver's station. Even if so, further development of the vision system capabilities and usage is very desired.