DIKU ITU KVL

Copenhagen Programming Language Seminar

COPLAS Talk

A Uniform Approach toward Nested Parallelism

Kazuhiko Kakehi,
University of Tokyo, Japan

Monday, September 6th, 15:15-16:00
DIKU, Universitetsparken 1, room N037

Abstract:

Parallel skeletons and homomorphisms over lists provide successful methods for parallelization. Several implementations exist due to their simplicity, which casts a contrast to more complicated data structures like trees. With assigning additional information properly, we know data structures can be coded into a list.

This paper develops a unified approach toward nested parallelism where data structures are represented into lists of tuples with depth information. Generic parallel computation schemes like reduce and accumulations are defined using the idea of list homomorphism. We demonstrate its uniformity and expressiveness using examples.

Scientific host: Robert Glück. Administrative host: Camilla Jensen. All are welcome.
The Copenhagen Programming Language Seminar (COPLAS) is a collaboration between DIKU, ITU and KVL.
To receive information about COPLAS talks by email, send a message to prog-lang-request@mail.it-c.dk with the word 'subscribe' as subject or in the body.

For more information about COPLAS, see http://www.coplas.org