Graduation Semester and Year

Spring 2024

Language

English

Document Type

Thesis

Degree Name

Master of Science in Computer Science

Department

Computer Science and Engineering

First Advisor

Dr. Leonidas Fegaras

Second Advisor

Dr. Zaman Noor

Third Advisor

Prof. David Levine

Abstract

This thesis introduces an innovative approach within the DIABLO (Data-Intensive Array-Based Loop Optimizer) framework, which integrates NumPy-inspired syntax and functionalities to facilitate the transition from single-node, array-based scientific computing to scalable, distributed data-parallel programming. This integration aims to reduce the learning curve and enhance the usability of distributed computing technologies for scientific researchers traditionally accustomed to NumPy’s operational paradigms. DIABLO leverages advanced distributed computing techniques to optimize traditional matrix operations, such as addition, subtraction, and multiplication, critical for numerous applications ranging from physics to machine learning. By reinterpreting these matrix operations to run efficiently over distributed architectures, DIABLO not only ensures computational integrity and scalability but also significantly enhances execution speeds compared to traditional single-node implementations. The effectiveness of DIABLO is demonstrated through detailed benchmarks that compare its performance with traditional methods, highlighting substantial improvements in computational efficiency and resource utilization. The results affirm that DIABLO’s approach to integrating familiar numerical computing techniques with robust distributed processing capabilities sets a new standard for scientific computing, making it an indispensable tool for researchers dealing with large-scale data sets.

Keywords

Diablo, Array, Numpy, Matrix, Matrix arithmetics

License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

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.