Title :
Approximate acceleration: A path through the era of dark silicon and big data
Author :
Hadi Esmaeilzadeh
Author_Institution :
Georgia Institute of Technology, School of Computer Science, Alternative Computing Technologies (ACT) Laboratory, United States
Abstract :
We have focused on two major research directions that leverage approximation to provide cross-stack solutions for (1) accelerating computation and (2) addressing the bottlenecks of data communication and storage.
Keywords :
"Approximation methods","Computational modeling","Acceleration","Big data","Accuracy","Approximation algorithms","Algorithm design and analysis"
Conference_Titel :
Compilers, Architecture and Synthesis for Embedded Systems (CASES), 2015 International Conference on
DOI :
10.1109/CASES.2015.7324540
