{"title":"RTL power optimisation: concepts, tools and design experiences [Tutorial]","authors":"M. Poncino, P. Sithambaram, R. Zafalon","doi":"10.1109/DATE.2004.1268816","DOIUrl":null,"url":null,"abstract":"The purpose of this master class is to present to the Digital System Design community a set of effective techniques for solving large scale combinatorial optimisation problems related to hardware and software co-design. In general, these problems are faced by modelling and solving them via Integer Programming (IP) techniques. Recently, Constraint Programming (CP) has emerged as a powerful programming paradigm that can be used in alternative or in conjunction with Integer Programming. Constraint Programming integrated concepts from different areas such as Artificial Intelligence, Mathematical Programming, Networks and Computational Logic. Its main strength concerns its efficiency, simplicity and flexibility. In particular flexibility is fundamental for changing the problem model adding or removing constraints without changing the solver. In the master class we 1) focus on finite domain Constraint Programming and its integration with Integer Programming, 2) describe system level design applications modelled via Constraint Programming 3) present ILOG, a leading edge, commercial tool embedding both Linear and Constraint Programming solvers. The objective of this master class is that of describing how emerging design methodologies for RTL power optimisation have found their way into commercial EDA tools, and how such tools have been successfully exploited in industry-strength designs. The course is organised into three main sections. The first one provides a review of the most effective RTL power optimisation techniques currently available. The second part is dedicated to the presentation and demonstration of innovative commercial EDA tools that implement the surveyed estimation and optimisation techniques. The third part reports on industrial experience on the usage of the methodologies and tools introduced in the previous sections. Intended audience for this class includes designers and design team managers from semiconductor companies and system houses, R&D engineers from EDA companies, and academic researchers and Ph.D. students in the field of IC/system design.","PeriodicalId":335658,"journal":{"name":"Proceedings Design, Automation and Test in Europe Conference and Exhibition","volume":"119 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2004-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings Design, Automation and Test in Europe Conference and Exhibition","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/DATE.2004.1268816","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
The purpose of this master class is to present to the Digital System Design community a set of effective techniques for solving large scale combinatorial optimisation problems related to hardware and software co-design. In general, these problems are faced by modelling and solving them via Integer Programming (IP) techniques. Recently, Constraint Programming (CP) has emerged as a powerful programming paradigm that can be used in alternative or in conjunction with Integer Programming. Constraint Programming integrated concepts from different areas such as Artificial Intelligence, Mathematical Programming, Networks and Computational Logic. Its main strength concerns its efficiency, simplicity and flexibility. In particular flexibility is fundamental for changing the problem model adding or removing constraints without changing the solver. In the master class we 1) focus on finite domain Constraint Programming and its integration with Integer Programming, 2) describe system level design applications modelled via Constraint Programming 3) present ILOG, a leading edge, commercial tool embedding both Linear and Constraint Programming solvers. The objective of this master class is that of describing how emerging design methodologies for RTL power optimisation have found their way into commercial EDA tools, and how such tools have been successfully exploited in industry-strength designs. The course is organised into three main sections. The first one provides a review of the most effective RTL power optimisation techniques currently available. The second part is dedicated to the presentation and demonstration of innovative commercial EDA tools that implement the surveyed estimation and optimisation techniques. The third part reports on industrial experience on the usage of the methodologies and tools introduced in the previous sections. Intended audience for this class includes designers and design team managers from semiconductor companies and system houses, R&D engineers from EDA companies, and academic researchers and Ph.D. students in the field of IC/system design.