This paper presents a comparison between superscalar and vector processors. First, we start with a detailed ISA analysis of the vector machine, including data related to masked execution, vector length and vector first facilities. Then we present a comparison of the two models at the instruction set architecture (ISA) level that shows that the vector model has several advantages: executes fewer instructions, fewer overall operations, and generally executes fewer memory accesses. We then analyse both models in terms of speculative execution, each one in its context. Results show that superscalar processors make an extensive use of speculation and that there is a large amount of misspeculated instructions. In the vector model, speculation is achieved using vector masks and, in general, fewer operations are misspeculated.

0,872

560

548

13

Q1