Parallelization and characterization of pattern matching using GPUs

Giorgos Vasiliadis, Michalis Polychronakis, Sotiris Ioannidis

Research output: Chapter in Book/Report/Conference proceedingConference contribution

19 Citations (Scopus)

Abstract

Pattern matching is a highly computationally intensive operation used in a plethora of applications. Unfortunately, due to the ever increasing storage capacity and link speeds, the amount of data that needs to be matched against a given set of patterns is growing rapidly. In this paper, we explore how the highly parallel computational capabilities of commodity graphics processing units (GPUs) can be exploited for high-speed pattern matching. We present the design, implementation, and evaluation of a pattern matching library running on the GPU, which can be used transparently by a wide range of applications to increase their overall performance. The library supports both string searching and regular expression matching on the NVIDIA CUDA architecture. We have also explored the performance impact of different types of memory hierarchies, and present solutions to alleviate memory congestion problems. The results of our performance evaluation using off-the-self graphics processors demonstrate that GPU-based pattern matching can reach tens of gigabits per second on different workloads.

Original languageEnglish
Title of host publicationProceedings - 2011 IEEE International Symposium on Workload Characterization, IISWC - 2011
Pages216-225
Number of pages10
DOIs
Publication statusPublished - 2011
Externally publishedYes
Event2011 IEEE International Symposium on Workload Characterization, IISWC - 2011 - Austin, TX, United States
Duration: 6 Nov 20118 Nov 2011

Other

Other2011 IEEE International Symposium on Workload Characterization, IISWC - 2011
CountryUnited States
CityAustin, TX
Period6/11/118/11/11

Fingerprint

Pattern matching
Data storage equipment
Graphics processing unit

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Computer Science Applications

Cite this

Vasiliadis, G., Polychronakis, M., & Ioannidis, S. (2011). Parallelization and characterization of pattern matching using GPUs. In Proceedings - 2011 IEEE International Symposium on Workload Characterization, IISWC - 2011 (pp. 216-225). [6114181] https://doi.org/10.1109/IISWC.2011.6114181

Parallelization and characterization of pattern matching using GPUs. / Vasiliadis, Giorgos; Polychronakis, Michalis; Ioannidis, Sotiris.

Proceedings - 2011 IEEE International Symposium on Workload Characterization, IISWC - 2011. 2011. p. 216-225 6114181.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Vasiliadis, G, Polychronakis, M & Ioannidis, S 2011, Parallelization and characterization of pattern matching using GPUs. in Proceedings - 2011 IEEE International Symposium on Workload Characterization, IISWC - 2011., 6114181, pp. 216-225, 2011 IEEE International Symposium on Workload Characterization, IISWC - 2011, Austin, TX, United States, 6/11/11. https://doi.org/10.1109/IISWC.2011.6114181
Vasiliadis G, Polychronakis M, Ioannidis S. Parallelization and characterization of pattern matching using GPUs. In Proceedings - 2011 IEEE International Symposium on Workload Characterization, IISWC - 2011. 2011. p. 216-225. 6114181 https://doi.org/10.1109/IISWC.2011.6114181
Vasiliadis, Giorgos ; Polychronakis, Michalis ; Ioannidis, Sotiris. / Parallelization and characterization of pattern matching using GPUs. Proceedings - 2011 IEEE International Symposium on Workload Characterization, IISWC - 2011. 2011. pp. 216-225
@inproceedings{ec706a74d042475f823b1783db983065,
title = "Parallelization and characterization of pattern matching using GPUs",
abstract = "Pattern matching is a highly computationally intensive operation used in a plethora of applications. Unfortunately, due to the ever increasing storage capacity and link speeds, the amount of data that needs to be matched against a given set of patterns is growing rapidly. In this paper, we explore how the highly parallel computational capabilities of commodity graphics processing units (GPUs) can be exploited for high-speed pattern matching. We present the design, implementation, and evaluation of a pattern matching library running on the GPU, which can be used transparently by a wide range of applications to increase their overall performance. The library supports both string searching and regular expression matching on the NVIDIA CUDA architecture. We have also explored the performance impact of different types of memory hierarchies, and present solutions to alleviate memory congestion problems. The results of our performance evaluation using off-the-self graphics processors demonstrate that GPU-based pattern matching can reach tens of gigabits per second on different workloads.",
author = "Giorgos Vasiliadis and Michalis Polychronakis and Sotiris Ioannidis",
year = "2011",
doi = "10.1109/IISWC.2011.6114181",
language = "English",
isbn = "9781457720642",
pages = "216--225",
booktitle = "Proceedings - 2011 IEEE International Symposium on Workload Characterization, IISWC - 2011",

}

TY - GEN

T1 - Parallelization and characterization of pattern matching using GPUs

AU - Vasiliadis, Giorgos

AU - Polychronakis, Michalis

AU - Ioannidis, Sotiris

PY - 2011

Y1 - 2011

N2 - Pattern matching is a highly computationally intensive operation used in a plethora of applications. Unfortunately, due to the ever increasing storage capacity and link speeds, the amount of data that needs to be matched against a given set of patterns is growing rapidly. In this paper, we explore how the highly parallel computational capabilities of commodity graphics processing units (GPUs) can be exploited for high-speed pattern matching. We present the design, implementation, and evaluation of a pattern matching library running on the GPU, which can be used transparently by a wide range of applications to increase their overall performance. The library supports both string searching and regular expression matching on the NVIDIA CUDA architecture. We have also explored the performance impact of different types of memory hierarchies, and present solutions to alleviate memory congestion problems. The results of our performance evaluation using off-the-self graphics processors demonstrate that GPU-based pattern matching can reach tens of gigabits per second on different workloads.

AB - Pattern matching is a highly computationally intensive operation used in a plethora of applications. Unfortunately, due to the ever increasing storage capacity and link speeds, the amount of data that needs to be matched against a given set of patterns is growing rapidly. In this paper, we explore how the highly parallel computational capabilities of commodity graphics processing units (GPUs) can be exploited for high-speed pattern matching. We present the design, implementation, and evaluation of a pattern matching library running on the GPU, which can be used transparently by a wide range of applications to increase their overall performance. The library supports both string searching and regular expression matching on the NVIDIA CUDA architecture. We have also explored the performance impact of different types of memory hierarchies, and present solutions to alleviate memory congestion problems. The results of our performance evaluation using off-the-self graphics processors demonstrate that GPU-based pattern matching can reach tens of gigabits per second on different workloads.

UR - http://www.scopus.com/inward/record.url?scp=84856182860&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84856182860&partnerID=8YFLogxK

U2 - 10.1109/IISWC.2011.6114181

DO - 10.1109/IISWC.2011.6114181

M3 - Conference contribution

SN - 9781457720642

SP - 216

EP - 225

BT - Proceedings - 2011 IEEE International Symposium on Workload Characterization, IISWC - 2011

ER -