Embedded Streaming Multimedia Computation with the RSVP(TM) Architecture

The need to process multimedia data places large computational demands on portable/embedded devices. These multimedia functions share common characteristics: they are computationally intensive and data-streaming, performing the same operation(s) on many data elements. The Reconfigurable Streaming Vector Processor (RSVP(TM)) is a vector coprocessor architecture that accelerates streaming data operations. Programming the RSVP(TM) architecture involves describing the shape and location of vector streams in memory and describing computations as data-flow graphs. These descriptions are intuitive and independent of each other, making the RSVP(TM) architecture easy to program, and they expose parallelism, enabling high-preformance from compiled code. They are also machine independent, allowing compatible implementations with varying cost-performance tradeoffs.

This talk will present the RSVP(TM) architecture and programming model, and our first two implementations. Our results show significant speedups on streaming data functions. Speedups for kernels and applications range from 2 to over 20 times that of a scalar host processor alone.