Opportunities for performance optimization of applications through code generation

Leiden Repository

Opportunities for performance optimization of applications through code generation

Title: Opportunities for performance optimization of applications through code generation
Author: Vu, Van Thieu
Publisher: Leiden Institute of Advanced Computer Science (LIACS), Faculty of Science, Leiden University
Issue Date: 2012-03-22
Keywords: Code generation
Optimizations
Parallel computing
GPU computing
Weather forecast model
Performance optimization
Finite element methods
Abstract: This thesis presents the new approaches to improve the performance of a weather forecast model. These approaches require the code complexities, which can be resolved by code generation. In Chapter 2 we investigated the extension to the code generation tool CTADEL to generate code for Galerkin finite element methods. Applying to generate code for the Shallow-Water equations, we found that the generated code is 3 and 1.2 times faster than the handwritten code with the gfortran 4.1.2 and pathscale 3.0 compiler, respectively. In Chapter 3, we optimized the parallel implementation of the HIRLAM weather forecast model by overlapping communications with calculations. We found that the performance of this parallel implementation can be significantly improved by overlapping communications with calculations. In Chapter 4 we extended CTADEL to generate parallel programs. By applying this technique we have successfully generated efficient parallel code for the Shallow-Water equations. In Chapter 5 we investigated a method to speed up the HIRLAM weather forecast model. We accelerated the dynamics routine by porting it on GPUs. We found that using GPUs for weather forecasting yields an order of magnitude performance improvement over the use of conventional CPUs. In Chapter 6 we showed our extension to CTADEL to automatically generate high efficient CUDA codes. Then, we applied to generate CUDA stream code for the dynamics routine of the HIRLAM weather forecast model. The results showed that the generated code is more efficient than the optimized handwritten program.
Description: Promotor: H.A.G. Wijshoff, Co-Promotor: A.A. Wolters
Faculty: Faculteit der Wiskunde en Natuurwetenschappen
Citation: Vu, V.T., 2012, Leiden University
ISBN: 9789088913907
Sponsor: NWO
Handle: http://hdl.handle.net/1887/18622
 

Files in this item

Description Size View
application/pdf Full Text 2.794Mb Under embargo
application/pdf Cover 1.453Mb View/Open
application/pdf Title Pages_Contents 1.692Mb View/Open
application/pdf Chapter 1 345.8Kb View/Open
application/pdf Chapter 2 533.5Kb View/Open
application/pdf Chapter 3 476.8Kb View/Open
application/pdf Chapter 4 453.0Kb View/Open
application/pdf Chapter 5 418.0Kb Under embargo
application/pdf Chapter 6 590.7Kb View/Open
application/pdf Chapter 7 158.8Kb View/Open
application/pdf Summary in Dutch 145.2Kb View/Open
application/pdf Acknowledgements_Curriculum Vitae_Bibliography 168.5Kb View/Open
application/pdf Propositions 86.07Kb View/Open

This item appears in the following Collection(s)