東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Kernel Specialization for Improved A...

Moore, Nicholas John.

Linked to FindBook

Google Book

Amazon

博客來

Kernel Specialization for Improved Adaptability and Performance on Graphics Processing Units (GPUs).

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Kernel Specialization for Improved Adaptability and Performance on Graphics Processing Units (GPUs)./
Author:	Moore, Nicholas John.
Description:	177 p.
Notes:	Source: Dissertation Abstracts International, Volume: 74-02(E), Section: B.
Contained By:	Dissertation Abstracts International74-02B(E).
Subject:	Engineering, Computer. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3527609
ISBN:	9781267620156

Kernel Specialization for Improved Adaptability and Performance on Graphics Processing Units (GPUs).
Moore, Nicholas John.

Kernel Specialization for Improved Adaptability and Performance on Graphics Processing Units (GPUs). - 177 p.

Source: Dissertation Abstracts International, Volume: 74-02(E), Section: B.

Thesis (Ph.D.)--Northeastern University, 2012.

Graphics processing units (GPUs) offer significant speedups over CPUs for certain classes of applications. However, maximizing GPU performance can be a difficult task due to the relatively high programming complexity as well as frequent hardware changes. Important performance optimizations are applied by the GPU compiler ahead of time and require fixed parameter values at compile time. As a result, many GPU codes offer minimum levels of adaptability to variations among problem instances and hardware configurations. These factors limit code reuse and the applicability of GPU computing to a wider variety of problems. This dissertation introduces GPGPU kernel specialization, a technique that can be used to describe highly adaptable kernels that work across different generations of GPUs with high performance. With kernel specialization, customized GPU kernels incorporating both problem- and implementation-specific parameters are compiled for each problem and hardware instance combination. This dissertation explores the implementation and parameterization of three real world applications targeting two generations of NVIDIA CUDA-enabled GPUs and utilizing kernel specialization: large template matching, particle image velocimetry, and cone-beam image reconstruction via backprojection. Starting with high performance adaptable GPU kernels that compare favorably to multi-threaded and FPGA-based reference implementations, kernel specialization is shown to maintain adaptability while providing performance improvements in terms of speedups and reduction in per-thread register usage. The proposed technique offers productivity benefits, the ability to adjust parameters that otherwise must be static, and a means to increase the complexity and parameterizability of GPGPU implementations beyond what would otherwise be feasible on current GPUs.

ISBN: 9781267620156Subjects--Topical Terms:

1669061
Engineering, Computer.

Kernel Specialization for Improved Adaptability and Performance on Graphics Processing Units (GPUs).
LDR:02888nam 2200301 4500 001 1957282
005 20131202131327.5
008 150210s2012 ||||||||||||||||| ||eng d
020 $a 9781267620156
035 $a (UMI)AAI3527609
035 $a AAI3527609
040 $a UMI $c UMI
100 1 $a Moore, Nicholas John. $3 2092149
245 1 0 $a Kernel Specialization for Improved Adaptability and Performance on Graphics Processing Units (GPUs).
300 $a 177 p.
500 $a Source: Dissertation Abstracts International, Volume: 74-02(E), Section: B.
500 $a Adviser: Miriam Leeser.
502 $a Thesis (Ph.D.)--Northeastern University, 2012.
520 $a Graphics processing units (GPUs) offer significant speedups over CPUs for certain classes of applications. However, maximizing GPU performance can be a difficult task due to the relatively high programming complexity as well as frequent hardware changes. Important performance optimizations are applied by the GPU compiler ahead of time and require fixed parameter values at compile time. As a result, many GPU codes offer minimum levels of adaptability to variations among problem instances and hardware configurations. These factors limit code reuse and the applicability of GPU computing to a wider variety of problems. This dissertation introduces GPGPU kernel specialization, a technique that can be used to describe highly adaptable kernels that work across different generations of GPUs with high performance. With kernel specialization, customized GPU kernels incorporating both problem- and implementation-specific parameters are compiled for each problem and hardware instance combination. This dissertation explores the implementation and parameterization of three real world applications targeting two generations of NVIDIA CUDA-enabled GPUs and utilizing kernel specialization: large template matching, particle image velocimetry, and cone-beam image reconstruction via backprojection. Starting with high performance adaptable GPU kernels that compare favorably to multi-threaded and FPGA-based reference implementations, kernel specialization is shown to maintain adaptability while providing performance improvements in terms of speedups and reduction in per-thread register usage. The proposed technique offers productivity benefits, the ability to adjust parameters that otherwise must be static, and a means to increase the complexity and parameterizability of GPGPU implementations beyond what would otherwise be feasible on current GPUs.
590 $a School code: 0160.
650 4 $a Engineering, Computer. $3 1669061
690 $a 0464
710 2 $a Northeastern University. $b Electrical and Computer Engineering. $3 1018491
773 0 $t Dissertation Abstracts International $g 74-02B(E).
790 1 0 $a Leeser, Miriam, $e advisor
790 1 0 $a Camps, Octavia $e committee member
790 1 0 $a Lebak, James $e committee member
790 1 0 $a Smith King, Laurie $e committee member
790 $a 0160
791 $a Ph.D.
792 $a 2012
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3527609