Abstract
Simulation studies of two branch prediction schemes in superscalar pipelined computers are presented in this paper. The two studied meth ods are the Per-address Adaptive (PA) scheme and the Global Adaptive (GA) scheme. In addi tion, simulation studies of a modification of the PA scheme, called Switched P/G Adaptive (SPGA), are also presented. Using several benchmarks, it has been illustrated that SPGA scheme has promising results at the possible expense of slight hardware addition.
Original language | English |
---|---|
Pages (from-to) | 227-242 |
Number of pages | 16 |
Journal | Simulation |
Volume | 74 |
Issue number | 4 |
DOIs | |
State | Published - Apr 2000 |
Keywords
- Computer architecture
- bench mark
- branch prediction
- performance