Abstract

Abstract
This paper discusses the design of YapDss, an or-parallel Prolog system for distributed memory parallel machines, such as the Beowulf PC clusters. The system builds on the work of YapOr, an or-parallel system for shared memory machines, and uses the distributed stack splitting binding model to represent computation state and work sharing among the computational workers. A new variant scheme of stack splitting, the diagonal splitting, is proposed and implemented. This scheme includes efficient algorithms to balance work load among computing workers, to determine the bottommost common node between two workers, and to calculate exactly the work load of one worker. An initial evaluation of the system shows that it is able to achieve very good speedups on a Beowulf PC cluster.

Abstract

Abstract
We describe a reference implementation of a multi-threaded run-time system for a core programming language based on a process calculus. The core language features processes running in parallel and communicating through asynchronous messages as the fundamental abstractions. The programming style is fully declarative, focusing on the interaction patterns between processes. The parallelism, implicit in the syntax of the programs, is effectively extracted by the language compiler and explored by the run-time system.

Abstract
Logic programming is based on the idea that computation is controlled inference. The Extended Andorra Model provides a very powerful framework that supports both co-routining and parallelism. In this work we show that David H. D. Warren's design for the EAM with Implicit Control does not perform well for deterministic computations and we present several optimisations that allow the BEAM to achieve performance matching or even exceeding related systems. Our optimisations refine the original EAM control rule demonstrate that overheads can be reduced through combined execution rules, and show that a good design and emulator implementation is relevant, even for a complex system such as the BEAM.

Abstract
This work aims at improving the scalability of memory usage in Inductive Logic Programming systems. In this context, we propose two efficient data structures: the Trie, used to represent lists and clauses; and the RL-Tree, a novel data structure used to represent the clauses coverage. We evaluate their performance in the April system using well known datasets. Initial results show a substantial reduction in memory usage without incurring extra execution time overheads. Our proposal is applicable in any ILP system.