Automatic data distribution optimisation in a lazy, self-optimising parallel matrix library (extended abstract)

Olav Beckmann and Paul H J Kelly

Department of Computing,
Imperial College, 180 Queen's Gate,
London SW7 2BZ, UK
Email: {ob3,phjk}@doc.ic.ac.uk

This short paper describes a matrix-vector library implementation running on the Fujitsu AP1000. The library optimises data distribution at run-time, taking advantage of information about how operands and results are used by delaying evaluation where possible. The work extends our earlier paper on the subject by giving a general methodology for representing data distributions, which is then used for formulating the optimisation problem and for describing an optimisation algorithm.