ThunderRW: An In-Memory Graph Random Walk Engine

Summary: ThunderRW is an in-memory graph random-walk engine enabling massive parallel walks with a step-centric Gather-Move-Update model and step interleaving to hide irregular memory access. It supports PPR, DeepWalk, Node2Vec, MetaPath, and outperforms prior systems by orders of magnitude, cutting CPU stalls from 73.1% to 15.0%. (summarized by gpt-5-nano on Feb 09 2026)

Paper ID: h17cab5a4e3934b47
Venue: VLDB
Year: 2021
Pagerank: 5.8677102e-05
Overall Rank: 6,108 | 59.06%
DOI: 10.14778/3476249.3476257
PDF: Download (CC BY-NC-ND 4.0)

Incoming Non-self Citations Over Time

Authors

1. Shixuan Sun (National University of Singapore)
2. Yuhang Chen (National University of Singapore)
3. Shengliang Lu (National University of Singapore)
4. Bingsheng He (National University of Singapore)
5. Yuchen Li (Singapore Management University)

BibTeX Citation

@article{sun_vldb21,
        title = {{ThunderRW: An In-Memory Graph Random Walk Engine}},
        author = {Sun, Shixuan and Chen, Yuhang and Lu, Shengliang and He, Bingsheng and Li, Yuchen},
        journal = {PVLDB},
        series = {{VLDB} '21},
        volume = {14},
        number = {11},
        pages = {1992--2005},
        doi = {10.14778/3476249.3476257},
        url = {https://doi.org/10.14778/3476249.3476257},
        year = {2021}
}

Incoming Citations (Sorted by Pagerank)

Showing 7 of 7 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
3,418	CoroGraph: Bridging Cache Efficiency and Work Efficiency for Graph Algorithm Execution	2024	VLDB	7.3032945e-05
7,279	An I/O-Efficient Disk-based Graph System for Scalable Second-Order Random Walk of Large Graphs	2022	VLDB	5.551099e-05
7,426	Space-Efficient Random Walks on Streaming Graphs	2023	VLDB	5.5145499e-05
9,193	LightRW: FPGA Accelerated Graph Dynamic Random Walks	2023	SIGMOD	5.1966793e-05
9,684	gSWORD: GPU-accelerated Sampling for Subgraph Counting	2024	SIGMOD	5.1258854e-05
10,173	uBlade: Efficient Batch Processing for Uncertain Graph Queries	2024	SIGMOD	5.0559291e-05
11,619	FlowWalker: A Memory-efficient and High-performance GPU-based Dynamic Graph Random Walk Framework	2024	VLDB	4.9605343e-05

Outgoing Citations (Sorted by Pagerank)

Showing 17 of 17 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Rank	Cited Paper	Year	Venue	Pagerank
3	Pregel: A System for Large-Scale Graph Processing	2010	SIGMOD	0.0012089097
22	Distributed GraphLab: A Framework for Machine Learning and Data Mining in the Cloud	2012	VLDB	0.00056028465
212	Sort vs. Hash Revisited: Fast Join Implementation on Modern Multi-Core CPUs	2009	VLDB	0.00024786576
548	Improving Index Performance through Prefetching	2001	SIGMOD	0.00016534019
902	Parallelizing Sequential Graph Computations	2017	SIGMOD	0.00013142001
931	Blogel: A Block-Centric Framework for Distributed Computation on Real-World Graphs	2014	VLDB	0.00012993132
2,008	GraphMat: High performance graph analytics made productive	2015	VLDB	9.1820424e-05
2,681	Asynchronous Memory Access Chaining	2016	VLDB	8.1203826e-05
2,814	Exploiting Coroutines to Attack the "Killer Nanoseconds"	2018	VLDB	7.9550699e-05
2,939	Interleaving with Coroutines: A Practical Approach for Robust Index Joins	2018	VLDB	7.82098e-05
3,470	Improving Main Memory Hash Joins on Intel Xeon Phi Processors: An Experimental Approach	2015	VLDB	7.2606844e-05
3,657	Realtime Top-k Personalized PageRank over Large Graphs on GPUs	2020	VLDB	7.1105244e-05
3,719	CoroBase: Coroutine-Oriented Main-Memory Database Engine	2021	VLDB	7.0525094e-05
4,097	TopPPR: Top-k Personalized PageRank Queries with Precision Guarantees on Large Graphs	2018	SIGMOD	6.7946571e-05
5,308	Parallel Personalized PageRank on Dynamic Graphs	2018	VLDB	6.1739397e-05
5,735	Cache-Efficient Fork-Processing Patterns on Large Graphs	2021	SIGMOD	5.9957538e-05
6,315	Memory-Aware Framework for Efficient Second-Order Random Walk on Large Graphs	2020	SIGMOD	5.7968933e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	4,052	READS: A Random Walk Approach for Efficient and Accurate Dynamic SimRank	2017	VLDB
2	2,103	BePI: Fast and Memory-Efficient Method for Billion-Scale Random Walk with Restart	2017	SIGMOD
3	10,889	ThunderGNN: Unlocking Tensor Cores for Graph Neural Networks	2026	VLDB
4	4,630	On the Embeddability of Random Walk Distances	2013	VLDB
5	8,041	Distributed Graph Embedding with Information-Oriented Random Walks	2023	VLDB
6	7,426	Space-Efficient Random Walks on Streaming Graphs	2023	VLDB
7	9,193	LightRW: FPGA Accelerated Graph Dynamic Random Walks	2023	SIGMOD
8	6,315	Memory-Aware Framework for Efficient Second-Order Random Walk on Large Graphs	2020	SIGMOD
9	7,279	An I/O-Efficient Disk-based Graph System for Scalable Second-Order Random Walk of Large Graphs	2022	VLDB
10	11,619	FlowWalker: A Memory-efficient and High-performance GPU-based Dynamic Graph Random Walk Framework	2024	VLDB