In this paper we study the parallel implementation of a traditional fr
ame based knowledge representation system for a general purpose massiv
ely parallel hypercube architecture (such as the Connection Machine CM
-2). We show that, using a widely available parallel system (instead o
f a special purpose architecture), it is possible to provide multiple
users with efficient shared access to a large scale knowledge-base. Pa
rallel algorithms are presented for answering multiple inference, asse
rt and retract queries on both single and multiple inheritance hierarc
hies. In addition to theoretical time complexity analysis, empirical r
esults obtained from extensive testing of a prototype implementation a
re presented.