登入選單
返回Google圖書搜尋
出版University of Tokyo, Faculty of Science, Department of Information Science, 1989
URLhttp://books.google.com.hk/books?id=DgX8GgAACAAJ&hl=&source=gbs_api
註釋Abstract: "Most of today's successful supercomputers have been developed in the restricted area of numerical computations featured by floating-point data processing. Conventional pipelining is still adopted to realize high-speed vector processing of linear equations since an array is a rigid data structure suited for the pipeline. From the point of evolution of application requiring more computational power, however, coming supercomputers should be used even for non-numeric, symbolic computations. We have to go toward using supercomputers for knowledge engineering and artificial intelligence applications in which non-numerical processing plays the leading part