An OR Parallel Prolog Model for Distributed Memory Systems

This paper shows a multisequential model to exploit OR parallelism on distributed memory systems. It presents an implementation of the incremental copy mechanism oriented to distributed systems and a novel distributed scheduler is also proposed. WAM modifications to support the proposed model are very simple. The system has been implemented on a 16 processor multicomputer based on transputers. It has been obtained very good performance results with an overhead around 6% and a speed-up comparable to the most known multisequential models for shared memory systems.

• Type: Conference
• Year: 1993
• Where: PLILP
• Authors: Vicente Benjumea, José M. Troya
• Keywords: Automated Reasoning, Distributed Memory, Memory Systems, Multisequential Models, PLILP 1993