The Stream Pram paper was written in 2005 and was co-written by Darrell Ulm.
This paper contained ideas about taking the abstract PRAM model, a module of parallel computation, and adding simultaneous multiple instruction streams for increased types of parallelism.
Here is a link: The Stream Pram Computer Science Paper by Darrell Ulm.
Some of the ideas here could apply to several actual parallel hardware, and the idea of SIMD applied parallel computation is applicable to modern GPU processors and gpGPUs which are excellent for tasks such as ray-tracing, and other graphics applications.
Clusters of GPUs can be thought of as multi-stream PRAM for the purpose of creating parallel algorithms.