Item Details

Print View

Superscalable Algorithms

Brunelle, Nathan
Format
Thesis/Dissertation; Online
Author
Brunelle, Nathan
Advisor
Robins, Gabriel
Abstract
We propose two new highly-scalable approaches to effectively process massive data sets in the post- Moore's Law era, namely (1) designing algorithms to operate directly on highly compressed data, and (2) leveraging massively parallel finite automata-based architectures for specific problem domains. The former method extends scalability by exploiting regularity in highly-compressible data, while also avoiding expensive decompression and re-compression. The latter hardware succinctly encapsulates complex behaviors via simulation of non-deterministic finite-state automata. We evaluate the efficiency, extensibility, and generality of these non-traditional approaches in big data environments. By presenting both promising experimental results and theoretical impossibility arguments, we provide more comprehensive frameworks for future research in these areas.
Language
English
Published
University of Virginia, Department of Computer Science, PHD (Doctor of Philosophy), 2017
Published Date
2017-08-05
Degree
PHD (Doctor of Philosophy)
Collection
Libra ETD Repository
Related Resources
https://www.youtube.com/watch?v=GP2rmOz3ebI http://www.cs.virginia.edu/~njb2b/dissertation/ http://www.cs.virginia.edu/~robins/compression/
Creative Commons Attribution LicenseCreative Commons Attribution License
▾See more
▴See less

Availability

Read Online