High-Performance Triangle Counting on GPUs

Hu, Yang; Liu, Huan; Huang, H. Howie

doi:10.1109/hpec.2018.8547570

Cited by 29 publications

(19 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In Table I, we show our performance in time and transacted edges per second (TEPS). We compare with last year's champion [10]. From the table we can see that we are consistently faster than Hu et al [10] in all real-world datasets, but slower for one synthetic dataset.…”

Section: B Resultsmentioning

confidence: 81%

“…We compare with last year's champion [10]. From the table we can see that we are consistently faster than Hu et al [10] in all real-world datasets, but slower for one synthetic dataset. Note that the given synthetic datasets have a larger number of triangles, which means more intermediate results will be generated during the computation and thus slow down our implementation.…”

Section: B Resultsmentioning

confidence: 81%

“…The total memory bandwidth of the Titan V is 652.8 GB/s. In our experiments, we compare with last year's champion [10]. In their paper, they run experiments on an NVIDIA Tesla P100 GPU from the San Diego Super Computer Center (SDSC).…”

Section: A Experimental Setupmentioning

confidence: 99%

“…The graphs are unlabeled graphs. Specifically, we compare results generated from small graphs recognized by Hu et al [10] for the reason that our implementation currently only supports one GPU and the GPU memory is limited. But our implementation could be extended to efficient multi-GPU implementation easily under the Gunrock framework.…”

Section: A Experimental Setupmentioning

confidence: 99%

See 3 more Smart Citations

Fast BFS-Based Triangle Counting on GPUs

Wang

Owens

2019

2019 IEEE High Performance Extreme Computing Conference (HPEC)

View full text Add to dashboard Cite

In this paper, we propose a novel method to compute triangle counting on GPUs. Unlike previous formulations of graph matching, our approach is BFS-based by traversing the graph in an all-source-BFS manner and thus can be mapped onto GPUs in a massively parallel fashion. Our implementation uses the Gunrock programming model and we evaluate our implementation in runtime and memory consumption compared with previous state-of-the-art work. We sustain a peak traversed-edgesper-second (TEPS) rate of nearly 10 GTEPS. Our algorithm is the most scalable and parallel among all existing GPU implementations and also outperforms all existing CPU distributed implementations. This work specifically focuses on leveraging our implementation on the triangle counting problem for the Subgraph Isomorphism Graph Challenge 2019, demonstrating a geometric mean speedup over the 2018 champion of 3.84×.

show abstract

Section: B Resultsmentioning

confidence: 81%

Section: B Resultsmentioning

confidence: 81%

Section: A Experimental Setupmentioning

confidence: 99%

Section: A Experimental Setupmentioning

confidence: 99%

See 2 more Smart Citations

Fast BFS-Based Triangle Counting on GPUs

Wang

Owens

2019

2019 IEEE High Performance Extreme Computing Conference (HPEC)

View full text Add to dashboard Cite

show abstract

“…The efforts of seeking suitable hardware platforms to accelerate triangle counting has also gain momentum. Popular attempts include multi-core CPUs [17], [18], [19], [20], manycore GPUs [21], [22], [23], [24], [25], and external memory devices [26], [27], [28], [29], [30]. Of all these platforms, GPUs are particularly tempting for the following reasons.…”

Section: Introductionmentioning

confidence: 99%

TRUST: Triangle Counting Reloaded on GPUs

Pandey,

Wang,

Zhong

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Triangle counting is a building block for a wide range of graph applications. Traditional wisdom suggests that i) hashing is not suitable for triangle counting, ii) edge-centric triangle counting beats vertex-centric design, and iii) communication-free and workload balanced graph partitioning is a grand challenge for triangle counting. On the contrary, we advocate that i) hashing can help the key operations for scalable triangle counting on Graphics Processing Units (GPUs), i.e., list intersection and graph partitioning, ii) vertex-centric design reduces both hash table construction cost and memory consumption, which is limited on GPUs. In addition, iii) we exploit graph and workload collaborative, and hashing-based 2D partitioning to scale vertex-centric triangle counting over 1,000 GPUs with sustained scalability. In this work, we present TRUST which performs triangle counting with the hash operation and vertex-centric mechanism at the core. To the best of our knowledge, TRUST is the first work that achieves over one trillion Traversed Edges Per Second (TEPS) rate for triangle counting.

show abstract

LPMA - An Efficient Data Structure for Dynamic Graph on GPUs

Zhang

Zou

2021

Web Information Systems Engineering – WISE 2021

View full text Add to dashboard Cite

High-Performance Triangle Counting on GPUs

Cited by 29 publications

References 9 publications

Fast BFS-Based Triangle Counting on GPUs

Fast BFS-Based Triangle Counting on GPUs

TRUST: Triangle Counting Reloaded on GPUs

LPMA - An Efficient Data Structure for Dynamic Graph on GPUs

Contact Info

Product

Resources

About