ABSTRACT
This chapter presents a brief sketch of the ILLIAC IV architecture and the nomenclature of the major architectural or organizational components which play a role in assessing the applicability of the ILLIAC IV architecture to image processing. It contains a description of several image processing applications on the ILLIAC IV together with a discussion of which architectural features of the machine are exploited in their formulation and which machine aspects create inefficiencies in the implementation. The chapter collects the various observations about the effectiveness of the ILLIAC IV architecture with regard to image processing. The computations required for clustering and classification are quite parallel, and thus well suited for ILLIAC IV use; processing turns out to be only a short job on the ILLIAC IV. The final station (FINST) hardware is divided into two sections, one dedicated to the processing of the overlap portion of the instruction, and the other dedicated to the execution portion of the instruction.
