ABSTRACT
Large-scale parallelism is the principal innovation to appear in the design of large commercially available computers in the 1980s. This chapter traces the history of this development and outlines the principles that can be applied to classify these parallel computers and characterise their relative performance. Individual designs may combine some or all of the parallel features. For example, a processor array may have pipelined arithmetic units as its pes, and one functional unit in a multi-unit computer might be a processor array. The main evolutionary trends and connections, evident in the introduction of parallelism into computer architecture. It is interesting to note that the serial organisation of arithmetic was considered as one of the advantages gained by introducing electronic components into computers, which had traditionally performed their arithmetic in parallel. The chapter also presents an overview of the key concepts discussed in this book.
