處理器的配置策略（processor allocation strategy），在一多處理機系統（multiprocessor system）上，扮演了一重要的角色。若能有效率的配置處理器，對整個多處理系統的效益（system performance），有莫大的助益。
在本篇論文中，我們針對超立方體多處理機（hypercube muliprocessor）系統，提出一快速且有效率的處理器配置策略。藉著重複地執行一些適當的移位折疊運算（shift folding operation），使得此法能夠辨識出一超立方體多處理機系統中所有子立方體（subcube）。
無論從理論上的分析或實驗的結果，都在在的顯示，我們的處理器配置策略，較以往所有已知的方法有更佳的系統效益。

When a task arrives at a hypercube multicomputer, it must be assigned to an availble subcube for execution. Efficient allocation of processors in the hypercube system is a key to its performance and utilization. In this thesis, we propose an efficient processor allocation scheme for hypercube multicomputer with lower allocation time comlexity than all previous strategies, while still achieving complete subcubes recognition. This scheme is based on repeatedly performing some appropriate shiftfolding operations on the allocation lists so that all possible subcubes of the desired dimension can be generated for recognition. The performance measures of different allocation strategies capable of achieving complete subcubes recognition are collected by conducting several independent simulation runs. From the analytic and experimental results, it can be demonstrated that our scheme has a better performance as compared with any other known strategy.