Developing Energy-Aware Strategies for the Blackfin Processor

Abstract

Energy usage is becoming an increasingly important design constraint for all computer systems. This issue is particularly critical in battery powered, embedded designs. Although many embedded processors have developed sophisticated power management schemes, few have produced an accurate, easy-to-use energy estimation framework. In this presentation we will describe the development of an instruction-level energy modeling framework for the Analog Devices Blackfin family of processors. Using this model, we are able to accurately estimate the energy consumed when running this code. While our main goal is to demonstrate that we can perform accurate energy estimation, we also plan to develop a framework that is fully integrated with compilation in order to produce more energy-efficient binaries. In this abstract we briefly describe our methodology and show data that illustrate some of the difficulties encountered when attempting to statically model energy.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 2005
Accession Number
ADA433304

Entities

People

  • David R. Kaeli
  • Giuseppe Olivadoti
  • Richard Gentile
  • Steven Vandersanden

Organizations

  • Northeastern University

Tags

DTIC Thesaurus Topics

  • Abstracts
  • Automated Speech Recognition
  • Computer Architecture
  • Computers
  • Computing System Architectures
  • Dissipation
  • Embedded Systems
  • Energy Consumption
  • Engineering
  • Families (Human)
  • Frequency
  • Image Processing
  • Instruction Set Architecture
  • Instructions
  • Measurement
  • Signal Processing
  • Standards

Fields of Study

  • Computer science

Readers

  • Distributed Systems and Data Platform Development
  • Electrical Engineering
  • Parallel and Distributed Computing.