Title :
A genetic approach to the design space exploration of superscalar microprocessor architectures
Author_Institution :
Rome Univ., Italy
Abstract :
This paper illustrates the results of a study on the automatic design space exploration of superscalar CPU architectures through genetic algorithms. The approach can take into account instruction level parallelism in conjunction with clock cycle time estimates, which could be a cumbersome task if it totally demanded human decisions
Keywords :
circuit CAD; genetic algorithms; integrated circuit design; logic CAD; microprocessor chips; clock cycle time estimates; design space exploration; genetic algorithms; instruction level parallelism; superscalar CPU architectures; superscalar microprocessor architectures; Algorithm design and analysis; Character generation; Clocks; Computer architecture; Genetic algorithms; Genetic mutations; Microarchitecture; Microprocessors; Sampling methods; Space exploration;
Conference_Titel :
Circuits and Systems, 2001. ISCAS 2001. The 2001 IEEE International Symposium on
Conference_Location :
Sydney, NSW
Print_ISBN :
0-7803-6685-9
DOI :
10.1109/ISCAS.2001.921987