High-Performance Massive Subgraph Counting Using Pipelined Adaptive-Group Communication. Chen, L., Peng, B., Ossen, S., Vullikanti, A., Marathe, M., Jiang, L., & Qiu, J. In Big Data and HPC: Ecosystem and Convergence, TopHPC 2017, pages 173-197, 2017. ![High-Performance Massive Subgraph Counting Using Pipelined Adaptive-Group Communication. [link]](https://bibbase.org/img/filetypes/link.svg "link")
Website doi abstract bibtex Subgraph counting aims to count the number of occurrences of a subgraph T (aka as a template) in a given graph G. The basic problem has found applications in diverse domains. The problem is known to be computationally challenging – the complexity grows both as a function of T and G. Recent applications have motivated solving such problems on massive networks with billions of vertices. In this chapter, we study the subgraph counting problem from a parallel computing perspective. We discuss efficient parallel algorithms for approximately resolving subgraph counting problems by using the color-coding technique. We then present several system-level strategies to substantially improve the overall performance of the algorithm in massive subgraph counting problems. We propose: 1) a novel pipelined Adaptive-Group communication pattern to improve inter-node scalability, 2) a fine-grained pipeline design to effectively reduce the memory space of intermediate results, 3) partitioning neighbor lists of subgraph vertices to achieve better thread concurrency and workload balance. Experimentation on an Intel Xeon E5 cluster shows that our implementation achieves 5x speedup of performance compared to the state-of-the-art work while reduces the peak memory utilization by a factor of 2 on large templates of 12 to 15 vertices and input graphs of 2 to 5 billions of edges.
@inproceedings{
title = {High-Performance Massive Subgraph Counting Using Pipelined Adaptive-Group Communication.},
type = {inproceedings},
year = {2017},
pages = {173-197},
websites = {https://doi.org/10.3233/978-1-61499-882-2-173},
id = {1ee48d86-fe5d-399b-8ecc-b8e41c276697},
created = {2019-10-01T17:21:01.407Z},
file_attached = {false},
profile_id = {42d295c0-0737-38d6-8b43-508cab6ea85d},
last_modified = {2020-05-11T14:43:31.543Z},
read = {false},
starred = {false},
authored = {true},
confirmed = {true},
hidden = {false},
citation_key = {Chen2017b},
source_type = {CPAPER},
private_publication = {false},
abstract = {Subgraph counting aims to count the number of occurrences of a subgraph T (aka as a template) in a given graph G. The basic problem has found applications in diverse domains. The problem is known to be computationally challenging – the complexity grows both as a function of T and G. Recent applications have motivated solving such problems on massive networks with billions of vertices. In this chapter, we study the subgraph counting problem from a parallel computing perspective. We discuss efficient parallel algorithms for approximately resolving subgraph counting problems by using the color-coding technique. We then present several system-level strategies to substantially improve the overall performance of the algorithm in massive subgraph counting problems. We propose: 1) a novel pipelined Adaptive-Group communication pattern to improve inter-node scalability, 2) a fine-grained pipeline design to effectively reduce the memory space of intermediate results, 3) partitioning neighbor lists of subgraph vertices to achieve better thread concurrency and workload balance. Experimentation on an Intel Xeon E5 cluster shows that our implementation achieves 5x speedup of performance compared to the state-of-the-art work while reduces the peak memory utilization by a factor of 2 on large templates of 12 to 15 vertices and input graphs of 2 to 5 billions of edges.},
bibtype = {inproceedings},
author = {Chen, Langshi and Peng, Bo and Ossen, Sabra and Vullikanti, Anil and Marathe, Madhav and Jiang, Lei and Qiu, Judy},
doi = {10.3233/978-1-61499-882-2-173},
booktitle = {Big Data and HPC: Ecosystem and Convergence, TopHPC 2017}
}
Downloads: 0
{"_id":"J3qi7QzRwCqNgzPLX","bibbaseid":"chen-peng-ossen-vullikanti-marathe-jiang-qiu-highperformancemassivesubgraphcountingusingpipelinedadaptivegroupcommunication-2017","authorIDs":[],"author_short":["Chen, L.","Peng, B.","Ossen, S.","Vullikanti, A.","Marathe, M.","Jiang, L.","Qiu, J."],"bibdata":{"title":"High-Performance Massive Subgraph Counting Using Pipelined Adaptive-Group Communication.","type":"inproceedings","year":"2017","pages":"173-197","websites":"https://doi.org/10.3233/978-1-61499-882-2-173","id":"1ee48d86-fe5d-399b-8ecc-b8e41c276697","created":"2019-10-01T17:21:01.407Z","file_attached":false,"profile_id":"42d295c0-0737-38d6-8b43-508cab6ea85d","last_modified":"2020-05-11T14:43:31.543Z","read":false,"starred":false,"authored":"true","confirmed":"true","hidden":false,"citation_key":"Chen2017b","source_type":"CPAPER","private_publication":false,"abstract":"Subgraph counting aims to count the number of occurrences of a subgraph T (aka as a template) in a given graph G. The basic problem has found applications in diverse domains. The problem is known to be computationally challenging – the complexity grows both as a function of T and G. Recent applications have motivated solving such problems on massive networks with billions of vertices. In this chapter, we study the subgraph counting problem from a parallel computing perspective. We discuss efficient parallel algorithms for approximately resolving subgraph counting problems by using the color-coding technique. We then present several system-level strategies to substantially improve the overall performance of the algorithm in massive subgraph counting problems. We propose: 1) a novel pipelined Adaptive-Group communication pattern to improve inter-node scalability, 2) a fine-grained pipeline design to effectively reduce the memory space of intermediate results, 3) partitioning neighbor lists of subgraph vertices to achieve better thread concurrency and workload balance. Experimentation on an Intel Xeon E5 cluster shows that our implementation achieves 5x speedup of performance compared to the state-of-the-art work while reduces the peak memory utilization by a factor of 2 on large templates of 12 to 15 vertices and input graphs of 2 to 5 billions of edges.","bibtype":"inproceedings","author":"Chen, Langshi and Peng, Bo and Ossen, Sabra and Vullikanti, Anil and Marathe, Madhav and Jiang, Lei and Qiu, Judy","doi":"10.3233/978-1-61499-882-2-173","booktitle":"Big Data and HPC: Ecosystem and Convergence, TopHPC 2017","bibtex":"@inproceedings{\n title = {High-Performance Massive Subgraph Counting Using Pipelined Adaptive-Group Communication.},\n type = {inproceedings},\n year = {2017},\n pages = {173-197},\n websites = {https://doi.org/10.3233/978-1-61499-882-2-173},\n id = {1ee48d86-fe5d-399b-8ecc-b8e41c276697},\n created = {2019-10-01T17:21:01.407Z},\n file_attached = {false},\n profile_id = {42d295c0-0737-38d6-8b43-508cab6ea85d},\n last_modified = {2020-05-11T14:43:31.543Z},\n read = {false},\n starred = {false},\n authored = {true},\n confirmed = {true},\n hidden = {false},\n citation_key = {Chen2017b},\n source_type = {CPAPER},\n private_publication = {false},\n abstract = {Subgraph counting aims to count the number of occurrences of a subgraph T (aka as a template) in a given graph G. The basic problem has found applications in diverse domains. The problem is known to be computationally challenging – the complexity grows both as a function of T and G. Recent applications have motivated solving such problems on massive networks with billions of vertices. In this chapter, we study the subgraph counting problem from a parallel computing perspective. We discuss efficient parallel algorithms for approximately resolving subgraph counting problems by using the color-coding technique. We then present several system-level strategies to substantially improve the overall performance of the algorithm in massive subgraph counting problems. We propose: 1) a novel pipelined Adaptive-Group communication pattern to improve inter-node scalability, 2) a fine-grained pipeline design to effectively reduce the memory space of intermediate results, 3) partitioning neighbor lists of subgraph vertices to achieve better thread concurrency and workload balance. Experimentation on an Intel Xeon E5 cluster shows that our implementation achieves 5x speedup of performance compared to the state-of-the-art work while reduces the peak memory utilization by a factor of 2 on large templates of 12 to 15 vertices and input graphs of 2 to 5 billions of edges.},\n bibtype = {inproceedings},\n author = {Chen, Langshi and Peng, Bo and Ossen, Sabra and Vullikanti, Anil and Marathe, Madhav and Jiang, Lei and Qiu, Judy},\n doi = {10.3233/978-1-61499-882-2-173},\n booktitle = {Big Data and HPC: Ecosystem and Convergence, TopHPC 2017}\n}","author_short":["Chen, L.","Peng, B.","Ossen, S.","Vullikanti, A.","Marathe, M.","Jiang, L.","Qiu, J."],"urls":{"Website":"https://doi.org/10.3233/978-1-61499-882-2-173"},"biburl":"https://bibbase.org/service/mendeley/42d295c0-0737-38d6-8b43-508cab6ea85d","bibbaseid":"chen-peng-ossen-vullikanti-marathe-jiang-qiu-highperformancemassivesubgraphcountingusingpipelinedadaptivegroupcommunication-2017","role":"author","metadata":{"authorlinks":{}},"downloads":0},"bibtype":"inproceedings","creationDate":"2019-08-21T14:56:43.703Z","downloads":0,"keywords":[],"search_terms":["high","performance","massive","subgraph","counting","using","pipelined","adaptive","group","communication","chen","peng","ossen","vullikanti","marathe","jiang","qiu"],"title":"High-Performance Massive Subgraph Counting Using Pipelined Adaptive-Group Communication.","year":2017,"biburl":"https://bibbase.org/service/mendeley/42d295c0-0737-38d6-8b43-508cab6ea85d","dataSources":["zgahneP4uAjKbudrQ","ya2CyA73rpZseyrZ8","2252seNhipfTmjEBQ"]}