Divide-and-conquer and parallel graph reduction
A transputer based parallel graph reduction system is described for the execution of functional programs following the divide-and-conquer- paradigm. The main idea is called dynamic partitioning, which refers to the maximum level of recursion up to which the system tries to solve subtasks in parallel and after which execution continues sequentially. Several experimental results on a mesh of 16 transputers are presented and discussed taking into account execution times, load balancing and task distribution. Based on the results obtained suggestions are made w.r.t. the partitioning depending on some characteristic properties of the programs.
- A parallel algorithm for finding minimum cutsets in reducible graphs
- N-graphs: Scalable topology and design of balanced divide-and-conquer algorithms
- Parallelization of divide-and-conquer in the Bird-Meertens formalism
- Pipelined functional tree accesses and updates: Scheduling, synchronization, caching and coherence
- A data-parallel formulation for divide and conquer algorithms
- scientific article; zbMATH DE number 3963189 (Why is no real title available?)
- Divide-and-conquer-based optimal parallel algorithms for some graph problems on EREW PRAM model
- Pipelines for Divide-and-Conquer Functions
This page was built for publication: Divide-and-conquer and parallel graph reduction
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1175201)
