Paper Info
Title
A Calculational Approach to Flattening Nested Data Parallelism in Functional Languages
Abstract
The data-parallel programming model is currently the most success- ful model for programming massively parallel computers. Unfortunately, it is, in its present form, restricted to exploiting flat data parallelism, which is not suitable for some classes of algorithms, e.g. those operating on irre gular structures. Re- cently, some effort has been made to implement nested data-parallel programs ef- ficiently by compiling them into equivalent flat programs usi ng a transformation called flattening. However, previous translations of nested into flat data-pa rallel programs have proved unwieldy when it comes to inventing and specifying opti- mizations and verifying the translation. This paper presen ts a new formalization of the flattening transformation in a calculational style. T he formalization is eas- ily verified and provides a good starting point for the develo pment of new opti- mizations. Some optimizations invented on the basis of this new formalism are described. Furthermore, we present practical evidence obt ained by experiment- ing with an implementation of the transformation.
Year
DOI
Venue
1996
10.1007/BFb0027796
ASIAN
Keywords
Field
DocType
functional languages,calculationa l method.,flattening transformation,implementation,calculational approach,functional programming,parallel programming,nested data parallelism,functional language,programming model
Programming language,Implicit parallelism,Programming paradigm,Task parallelism,Computer science,Parallel computing,Inductive programming,Parallel programming model,Data parallelism,Reactive programming,Declarative programming
Conference
ISBN
Citations 
PageRank 
3-540-62031-1
5
0.57
References 
Authors
8
2
Name
Order
Citations
PageRank
Gabriele Keller165736.02
Martin Simons2629.79