Date of Award

2017

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Electrical and Computer Engineering

Committee Chair

B. Earl Wells

Committee Member

Ken-Ichi Nishikawa

Committee Member

Aleksandar Milenkovic

Committee Member

Guo-Hui Zhang

Committee Member

Murat M. Tanik

Committee Member

Seong-Moo Yoo

Committee Member

Ravi Gorur

Subject(s)

Gate array circuits, Programmable array logic, Programmable logic devices, Space plasmas, Plasma (Ionized gases)--Mathematical models

Abstract

The primary focus of this research has been to expand and validate a framework through which Field Programmable Gate Array (FPGA) technology can be effectively utilized to reduce both the execution time and the energy consumption attributes of Particle-in-Cell (PIC) based simulations of space plasmas as compared to traditional single-core central processing unit-based implementations. This framework utilizes the OpenCL standard in a manner that allows the physical model to be captured at a high level of abstraction and allows the FPGA synthesis/compilation tool to exploit a combination of task and data parallelism. This research introduces an optimized implementation of the most time-consuming phase in the PIC simulation that is the particle to grid interpolation phase using a task parallel model. Compared to the naive and the optimized data parallel implementation models, this implementation offers up to 64 fold speed up and 80 fold speedup over a naive implementation. The framework has been shown to have attributes that scale well to future spatially reconfigurable architectures and can be further expanded to other particle to grid-based computational science applications. Simulations produced by this framework have demonstrated a nominal 2.5-fold improvement in performance and an eight-fold improvement in energy consumption over the life of the simulation when compared to the reference single core CPU implementation. The framework was verified and validated by comparing its results with existing 2.5 dimensional electrostatic PIC simulations that are being used to study the dynamical development of potential structures within the auroral regions of the Earth.

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.