Accelerating Triangle Counting on GPU

被引:17
作者
Hu, Lin [1 ]
Zou, Lei [1 ]
Liu, Yu [1 ]
机构
[1] Peking Univ, Beijing, Peoples R China
来源
SIGMOD '21: PROCEEDINGS OF THE 2021 INTERNATIONAL CONFERENCE ON MANAGEMENT OF DATA | 2021年
关键词
Triangle counting; GPU; Edge directing; Vertex ordering; DECOMPOSITION; NETWORKS; GRAPHS; MODEL;
D O I
10.1145/3448016.3452815
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Triangle counting is an important problem in graph mining, which has achieved great performance improvement on GPU in recent years. Instead of proposing a new GPU triangle counting algorithm, in this paper, we propose a novel lightweight graph preprocessing method to boost many state-of-the-art GPU triangle counting algorithms without changing their implementations and data structures. Specifically, we find common computing patterns in existing algorithms, and abstract two analytic models to measure how workload imbalance and diversity in these computing patterns affect performance exactly. Then, due to the NP-hardness of the model optimization, we propose approximate solutions by determining edge directions to balance workloads and reordering vertices to maximize the degree of parallelism within GPU blocks. Finally, extensive experiments confirm the significant performance improvement and high usability of our approach.
引用
收藏
页码:736 / 748
页数:13
相关论文
共 37 条
  • [1] Aiello W., 2000, Proceedings of the Thirty Second Annual ACM Symposium on Theory of Computing, P171, DOI 10.1145/335305.335326
  • [2] Finding and counting given length cycles
    Alon, N
    Yuster, R
    Zwick, U
    [J]. ALGORITHMICA, 1997, 17 (03) : 209 - 223
  • [3] A Simple BSP-based Model to Predict Execution Time in GPU Applications
    Amaris, Marcos
    Cordeiro, Daniel
    Goldman, Alfredo
    de Camargo, Raphael Y.
    [J]. 2015 IEEE 22ND INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING (HIPC), 2015, : 285 - 294
  • [4] [Anonymous], 2018, P 2018 IEEE HIGH PER
  • [5] [Anonymous], 2016, SIGPLAN NOTICES
  • [6] Ao NY, 2011, PROC VLDB ENDOW, V4, P470
  • [7] PATRIC: A Parallel Algorithm for Counting Triangles in Massive Networks
    Arifuzzaman, Shaikh
    Khan, Maleq
    Marathe, Madhav
    [J]. PROCEEDINGS OF THE 22ND ACM INTERNATIONAL CONFERENCE ON INFORMATION & KNOWLEDGE MANAGEMENT (CIKM'13), 2013, : 529 - 538
  • [8] Parallel Triangle Counting and Enumeration using Matrix Algebra
    Azad, Ariful
    Buluc, Aydin
    Gilbert, John
    [J]. 2015 IEEE 29TH INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM WORKSHOPS, 2015, : 804 - 811
  • [9] A subquadratic triad census algorithm for large sparse networks with small maximum degree
    Batagelj, V
    Mrvar, A
    [J]. SOCIAL NETWORKS, 2001, 23 (03) : 237 - 243
  • [10] High Performance Exact Triangle Counting on GPUs
    Bisson, Mauro
    Fatica, Massimiliano
    [J]. IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, 2017, 28 (12) : 3501 - 3510