Electronic Theses and Dissertations

Date of Award

2015

Document Type

Thesis

Degree Name

M.S. in Engineering Science

Department

Computer and Information Science

First Advisor

Byunghyun Jang

Second Advisor

Philip J. Rhodes

Third Advisor

Dawn Wilkins

Relational Format

dissertation/thesis

Abstract

As General Purpose GPUs (GPGPU) are increasingly becoming a prominent component of high performance computing platforms, power and thermal dissipation are getting more attention. The trade-offs among performance, power, and heat must be well modeled and evaluated from the early stage of GPU design. This necessitates a tool that allows GPU architects to quickly and accurately evaluate their design. There are a few models for GPU power but most of them estimate power at a higher level than architecture, which are therefore missing hardware reconfigurability. In this thesis, we propose a framework that models power and heat dissipation at the hardware architecture level, which allows for configuring and investigating individual hardware components. Our framework is also capable of visualizing the heat map of the processor over different clock cycles. To the best of our knowledge, this is the first comprehensive framework that integrates and visualizes power consumption and heat dissipation of GPUs.

Concentration/Emphasis

Emphasis: Computer Science

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.