Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

This dissertation by Nicolas Melot investigates energy-efficient parallel stream computing on many-core architectures. The work focuses on optimizing execution speed and energy consumption through dataflow computing and pipelined execution of streaming applications. It introduces Drake, a tool for generating and testing pipelined applications on many-core architectures, and discusses methodologies for optimizing energy efficiency given a fixed throughput target. The research includes case studies using Intel SCC and Xeon processors and aims to advance the development and optimization of parallel processing techniques to address the growing demands of big data computing.

Parallel processing (Electronic computers) --- Data flow computing.