A data-parallel language and its compilation to a formally defined intermediate language

Is it possible to design a language based on concepts well known to most programmers that can be used as a tool for efficient parallel programming? The goal is to use a basically serial language in which inherent parallelism is expressed precisely enough for a compiler to generate very efficient code. The authors believe that a good starting point is the data parallel programming style, which heavily exploits parallelism but still is deterministic. Many algorithms can be expressed in a sequential style where code acts on a collection of objects, such as a set, an array or a relation. Languages with data parallel constructs are often referred to as collection-oriented. This family contains a large number of languages, e.g., C*, and Paralation Lisp. This paper presents a new collection-oriented language. The authors work has also been influenced by the language Booster, and research on very high level languages such as Refine.<
>