Title: PHP: Power hungry pattern generation at higher abstraction level
Abstract: The Performance, area, and power are most essential factors to be considered and optimize at every step in the design cycle. Design engineers often need to learn about these factors in order make right decisions on design strategies. Power analysis at lower levels of abstraction can provide more accurate analysis than higher levels. Worst case power can be estimated through high activity pattern generation. However, generation of such power hungry patterns (PHP) become challenging as the number of modules or design components increases. This process can be accelerated at higher abstraction levels by utilizing the available information. In this paper, we generate PHP by at higher abstraction level with significant speed up. A genetic algorithm is implemented to find out the global maximum power of designs. An experiment indicates that the process implemented is much faster and it finds about 10 % more power demanding PHP than random samples generated.
Publication Year: 2017
Publication Date: 2017-09-01
Language: en
Type: article
Indexed In: ['crossref']
Access and Citation
Cited By Count: 3
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot