Efficient Lazy Data-Structures on a Dataflow Machine
MASSACHUSETTS INST OF TECH CAMBRIDGE LAB FOR COMPUTER SCIENCE
Pagination or Media Count:
Eager interpreters are able to exploit vast parallelism, yet lazy interpreters have more desirable termination properties. We propose lazy data-structures, an extension to the dataflow language Id, to support a combination of eager and lazy evaluation. We describe the semantics of lazy data-structures, as well as efficient implementation on the Tagged-Token Dataflow Architecture and the Monsoon Explicit Token Store Machine. We develop support for lazy data-structures in the language, the compiler, the run-time system, the interpreter, and the proposed hardware and demonstrate the effectiveness of the construct as well as the limitations. Keywords Dataflow, Evaluation order, Functional languages, Lazy evaluation, Parallel data-structures, Computer languages, Computer programming, Data processing.
- Computer Programming and Software