Title :
Energy Efficiency through Significance-Based Computing
Author :
Nikolopoulos, Dimitrios S. ; Vandierendonck, Hans ; Bellas, Nikolaos ; Antonopoulos, Christos D. ; Lalis, Spyros ; Karakonstantis, Georgios ; Burg, Andreas ; Naumann, Uwe
Abstract :
An extension of approximate computing, significance-based computing exploits applications´ inherent error resiliency and offers a new structural paradigm that strategically relaxes full computational precision to provide significant energy savings with minimal performance degradation.
Keywords :
energy conservation; energy consumption; fault tolerant computing; green computing; power aware computing; approximate computing; computational precision; energy efficiency; energy savings; error resiliency; performance degradation; significance-based computing; structural paradigm; Algorithm design and analysis; Circuit faults; Computers; Convergence; Educational institutions; Hardware; Reliability; Green computing; SCoRPiO; Significance-Based Computing for Reliability and Power Optimization; approximate computing; automatic differentiation; energy wall; hardware; iterative solvers; on-chip power consumption; significance-based computing; software engineering;
DOI :
10.1109/MC.2014.182
