Type

Text

Type

Thesis

Advisor

Hong, Sangjin | Doboli, Alex.

Date

2014-12-01

Keywords

BBDF, FPGA, Reconfigurable, SystemC | Engineering

Department

Department of Electrical Engineering.

Language

en_US

Source

This work is sponsored by the Stony Brook University Graduate School in compliance with the requirements for completion of degree.

Identifier

http://hdl.handle.net/11401/77486

Publisher

The Graduate School, Stony Brook University: Stony Brook, NY.

Format

application/pdf

Abstract

Modern reconfigurable logic devices are often multi-core and multi-processor architectures with complex intra-processor logic. To fully utilize the raw power of these devices is an arduous task; mapping design data-flows to such devices in a way that will maximize their performance involves careful consideration of a number of parameters, and is the subject of a good amount of research. This thesis presents a novel design, which uses the technique of buffer-based dataflow, a representation technique for realizing data-centric applications in reconfigurable platforms, to map complex logic systems with multiple processing elements to a reconfigurable target architecture having multi-core processors or multiple processors. The use of multi-core processors requires careful synchronization between the processing elements and we propose employing the buffer-based dataflow technique in conjunction with a controller to map the processing logic onto the reconfigurable platform and deal with the synchronization issues. The logic is implemented using a series of buffers and interconnections, and these are controlled by a top-level global controller, responsible for their configuration and reconfiguration as well as path selection to enable dynamic switching between designs. The dynamic reconfigurability gained from our approach allows us to map multiple processing elements onto a single core and switch between them during run-time while maximizing performance. The proposed design is evaluated with SystemC and Xilinx ISE. | 67 pages

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.