Functional Reactive Programming for Embedded Systems with GPGPUs.

We propose new functionality for FRP language to support computations on GPUs. Some embedded systems equip GPUs, which enable us to process large amounts of data such as camera images. We have developed an FRP language for embedded systems, named Emfrp. However, Emfrp lacks two kinds of concepts for computations on GPUs; one is the representation for large amounts of data, and the other is the declaration of the calculational resources of computations. In this paper, we introduce node arrays and GPU nodes, which are time-varying values computed on GPUs. We also show a prototype language of these features and evaluation through some examples on it.