|
In this section we present a brief overview of digital computer architectures, before getting into the description of optical computing architectures.
In the previous chapter we showed a very general computer architecture, which includes an input and an output, a memory for temporary or long-term storage, a processor or processor array to carry out its operations, and interconnections to tie all the parts together, and one possible classification scheme for computers. This is probably the most useful classification scheme for classifying optical computers. However for purposes of comparison we shall discuss other classification schemes used for digital computers.
The classification of parallel computer architectures is a difficult question because (1) most computers now have some degree of low-level parallelism; (2) the t.axonomy that was proposed a few years ago for classifying parallel computer architectures (Flynn's t.axonomy) has turned out to be insufficient: there are computer architectures that do not fit well into these schemes, or fitting them in is rather artificial; (3) although other classification schemes have been proposed, none so far has proven to be satisfactory.
Computers may be classified according to complexity, according to instruction and data streams, or according to type of problem.
|
