NESL

Carnegie Mellon

A fine-grained, functional, data-parallel language with nested data structures and nested parallelism. Includes a built-in parallel data type and parallel operations on sequences. Loosely based on ML[2]. Useful for parallel algorithms on sparse matrices and graphs.

"NESL: A Nested Data-Parallel Language", Guy Blelloch, CMU-CS-93-129, April 1993.