An important objective of the MicroUnity mediaprocessor is to allow the design of systems that replace hardwired functionality with software. One of the key design techniques that has been employed at the architectural level is an instruction set that processes 128-bit vectors of data of varying widths. This introduces the challenge of writing software that effectively uses the available vector processing across the broad range of algorithms of interest. In this paper, we focus on a selected set of algorithms that illustrate many of the techniques we have developed for vector parallel programming on the mediaprocessor.