Thesis
Evaluation of dynamic voltage and frequency scaling techniques for designing energy efficient multi-core systems
Washington State University
Master of Science (MS), Washington State University
2012
Handle:
https://hdl.handle.net/2376/101292
Abstract
High-Performance Computing is moving towards chip multi-processors, where in a single chip will have multiple processors. Placing a large number of processors on a single chip causes high energy consumption which causes cost, reliability and scalability issues. The high energy demand for these systems will also have an adverse impact on the environment. This calls for an investigation of how we can reduce the energy consumption in multi-core systems. The issue of energy consumption has been studied in the single processor domain. The processor will not be busy the whole time during the execution of an application. It will have idle periods interspersed between the busy period, which presents us an opportunity save energy. Input/output operations, memory fetch, and communication with other processors are some of the reasons why a processor will be idle during the task execution. Most of the methodologies studied focus on varying v the voltage/frequency during the idle period of the processors. These methodologies are called dynamic voltage and frequency scaling methodologies. The effects of dynamic voltage and frequency scaling on multi-core processors presents a challenging study. We explore the behavior and energy-performance trade-offs of the multi-core systems while running various applications when dynamic voltage and frequency techniques are employed. This gives us a better understanding of how the behavior of the multi-core system changes with the different workloads and helps in reducing the power consumption in emerging technology.
Metrics
9 File views/ downloads
11 Record Views
Details
- Title
- Evaluation of dynamic voltage and frequency scaling techniques for designing energy efficient multi-core systems
- Creators
- Rajath Kumar Shantharam Hegde
- Contributors
- Behrooz A. Shirazi (Degree Supervisor)
- Awarding Institution
- Washington State University
- Academic Unit
- Electrical Engineering and Computer Science, School of
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University; Pullman, Wash. :
- Identifiers
- 99900525079401842
- Language
- English
- Resource Type
- Thesis