Parka on Mimd-supercomputers
نویسندگان
چکیده
We have ported the SIMD Parka knowledge representation system to generic MIMD machines. The system has been recoded in C and supported using runtime optimization packages developed in the High Performance Systems Software Laboratory at the University of Maryland. New \scanning" algorithms have been developed for inheritance and recognition inferences. These algorithms have been tested with both random networks and on a recoding of the ontology of the CYC knowledge base as well as on large planning case-bases. Tests show that the new version is signiicantly faster than the SIMD system, and that it promises to scale well to knowledge bases orders of magnitude larger than CYC.
منابع مشابه
Utilising Mimd Parallelism in Modular Visualisation Environments
Modular Visualisation Environments such as apE 1, 2, 3, 4] or Explorer 6, 7] are increasingly being used in the visualisation community. Such systems can already place compute intensive modules on supercomputers in order to utilise their power. This paper describes the problems that arise in utilising distributed memory MIMD computers within this environment. The work presented here is driven b...
متن کاملPARKA: A System for Massively Parallel Knowledge Representation
PARKA is a frame-based knowledge representation system implemented on massively parallel hardware--the Connection Machine (CM-2). PARKA provides a representation language consisting of concept descriptions (frames) and binary relations on those descriptions (slots). The system is designed explicitly to provide extremely fast property inheritance inference capabilities. In particular, PARKA can ...
متن کاملA Comparison between Domain Decomposition and Fully Implicit Approaches for a Parallel 3d Upwind Flow Solver
An implicit method for the solution of the 3D supersonic Euler equations is presented. The numerical technique adopted is a nite volume method based on an unfactored implicit upwind scheme. A great relevance is given to the deenition of the implicit operator and to the solution of the linear systems. A preconditioned iterative solver has been used showing better performance than unpreconditione...
متن کاملFinal Report on Research in Parallel Computing : Network - based Multicomputers 1990 - 1995
No existing parallel architecture (including SIMD machines, MIMD machines, vector processors, machines with a common, globally-addressable memory, machines with private memories) has demonstrated its superiority for all possible applications. That is, given different applications, different parallel machines may deliver the highest absolute performance or the best price-performance. A good solu...
متن کاملThe Evolution of Anupam Supercomputers
Computers have become ubiquitous in all walks of scientific endeavour. There have been innumerable architectures of computers, aiding equally countless, diverse fields of research, thereby rendering the task of classification of computers itself a nontrivial job. Interestingly, this gamut of computers are classified neither by their explicit features nor by the class of application to which the...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 1995