Abstract: Along with the perfection of parallel computing algorithms and the maturity of cheap but powerful multiprocessing systems, parallel multi dimensional join and aggregation on data warehouse with the multiprocessing system becomes the preferred technology for efficiently improving OLAP operation In this paper, a new method（PJAMDDC algorithm） is proposed for processing time consuming join and aggregate operation In this method, by taking into consideration the characteristics of the storage mechanism of multi dimensional data cube and the structural characteristic of multiprocessing system and adopting hierarchy combining surrogate/search lattice of data cube with weight on the basis of the former logic structure of search lattice of data cube It advances the join and aggregate efficiency by achieving the optimal state for distribution of data cube between processors As illustrated by experimental result of the performance analysis, PJAMDDC algorithm is efficient when it is used for parallel multi dimensional join and aggregation on data warehouse