We take sorting of large data sets as case study for making data-intensive applications more energy-efficient. Using a low-power processor, solid state disks, and efficient algorithms, we beat the current records in the JouleSort benchmark for 10 GB to 1 TB of data by factors of up to 5.1. Since we also use parallel processing, this usually comes without a performance penalty.
Published in:
Green Computing Conference, 2010 International
Date of Conference: 15-18 Aug. 2010