MG-Join: A Scalable Join for Massively Parallel Multi-GPU Architectures

Summary: MG-Join proposes a scalable partitioned hash join for multi-GPU single-machine architectures. Adaptive multi-hop cross-GPU routing minimizes congestion, achieving up to 97% bisection-bandwidth utilization, and up to 2.5x join speedups with 4.5x TPC-H gains over Omnisci. (summarized by gpt-5-nano on Feb 09 2026)

Paper ID: 6207
Venue: SIGMOD
Year: 2021
Pagerank: 7.6217889e-05
Overall Rank: 3,227 | 77.87%
DOI: 10.1145/3448016.3457254

Incoming Non-self Citations Over Time

Authors

1. Johns Paul (National University of Singapore)
2. Shengliang Lu (National University of Singapore)
3. Bingsheng He (National University of Singapore)
4. Chiew Tong Lau (Nanyang Technological University)

BibTeX Citation

@inproceedings{paul_sigmod21,
        title = {{MG-Join: A Scalable Join for Massively Parallel Multi-GPU Architectures}},
        author = {Paul, Johns and Lu, Shengliang and He, Bingsheng and Lau, Chiew Tong},
        series = {{SIGMOD} '21},
        booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
        publisher = {Association for Computing Machinery},
        doi = {10.1145/3448016.3457254},
        url = {https://dl.acm.org/doi/10.1145/3448016.3457254},
        year = {2021}
}

Incoming Citations (Sorted by Pagerank)

Showing 20 of 20 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
4,144	Tile-based Lightweight Integer Compression in GPU	2022	SIGMOD	6.8744592e-05
4,243	Orchestrating Data Placement and Query Execution in Heterogeneous CPU-GPU DBMS	2022	VLDB	6.8073248e-05
4,535	Triton Join: Efficiently Scaling to a Large Join State on GPUs with Fast Interconnects	2022	SIGMOD	6.6419266e-05
5,053	RTIndex: Exploiting Hardware-Accelerated GPU Raytracing for Database Indexing	2023	VLDB	6.3844953e-05
5,422	GPU Database Systems Characterization and Optimization	2024	VLDB	6.2245373e-05
5,587	Distributed GPU Joins on Fast RDMA-capable Networks	2023	SIGMOD	6.1596139e-05
6,065	Vortex: Overcoming Memory Capacity Limitations in GPU-Accelerated Large-Scale Data Analytics	2025	VLDB	5.9895288e-05
6,774	Evaluating Multi-GPU Sorting with Modern Interconnects	2022	SIGMOD	5.7778738e-05
6,957	Design and Analysis of a Processing-in-DIMM Join Algorithm: A Case Study with UPMEM DIMMs	2023	SIGMOD	5.7303405e-05
7,058	BOSS - An Architecture for Database Kernel Composition	2024	VLDB	5.7154673e-05
7,432	Powerful GPUs or Fast Interconnects: Analyzing Relational Workloads on Modern GPUs	2025	VLDB	5.6199783e-05
7,591	Efficiently Processing Joins and Grouped Aggregations on GPUs	2025	SIGMOD	5.5900315e-05
7,959	Terabyte-Scale Analytics in the Blink of an Eye	2026	VLDB	5.5181056e-05
8,808	Scaling your Hybrid CPU-GPU DBMS to Multiple GPUs	2024	VLDB	5.3661351e-05
9,333	Efficiently Joining Large Relations on Multi-GPU Systems	2025	VLDB	5.2887551e-05
10,541	Scalable GPU Acceleration of Scalar Functions in Analytical Databases: Compilation, Benchmarking, and Optimization	2026	VLDB	5.093636e-05
10,985	Scaling GPU-Accelerated Databases beyond GPU Memory Size	2025	VLDB	5.093636e-05
11,204	SPID-Join: A Skew-resistant Processing-in-DIMM Join Algorithm Exploiting the Bank- and Rank-level Parallelisms of DIMMs	2024	SIGMOD	5.093636e-05
11,231	Accelerating Merkle Patricia Trie with GPU	2024	VLDB	5.093636e-05
11,556	Scaling Equi-Joins	2022	SIGMOD	5.093636e-05

Outgoing Citations (Sorted by Pagerank)

Showing 15 of 15 cited papers.

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

Rank	Cited Paper	Year	Venue	Pagerank
631	Relational Joins on Graphics Processors	2008	SIGMOD	0.00015591241
823	The Yin and Yang of Processing Data Warehousing Queries on GPU Devices	2013	VLDB	0.00013792901
892	Rack-Scale In-Memory Join Processing using RDMA	2015	SIGMOD	0.00013376761
1,265	An Experimental Comparison of Thirteen Relational Equi-Joins in Main Memory	2016	SIGMOD	0.00011415709
1,974	Track Join: Distributed Joins with Minimal Network Traffic	2014	SIGMOD	9.3658402e-05
1,977	HetExchange: Encapsulating heterogeneous CPU-GPU parallelism in JIT compiled engines	2019	VLDB	9.3641101e-05
2,140	Revisiting Co-Processing for Hash Joins on the Coupled CPU-GPU Architecture	2013	VLDB	9.0991487e-05
3,303	A Distributed Multi-GPU System for Fast Graph Processing	2018	VLDB	7.5405596e-05
3,471	In-Cache Query Co-Processing on Coupled CPU-GPU Architectures	2015	VLDB	7.3885861e-05
3,782	CROSSBOW: Scaling Deep Learning with Small Batch Sizes on Multi-GPU Servers	2019	VLDB	7.1298979e-05
3,791	Hardware-conscious Query Processing in GPU-accelerated Analytical Engines	2019	CIDR	7.1235328e-05
4,585	Data-Parallel Query Processing on Non-Uniform Data	2020	VLDB	6.6175897e-05
5,092	Ocelot/HyPE: Optimized Data Processing on Heterogeneous Hardware	2014	VLDB	6.3669454e-05
5,807	Improving Execution Efficiency of Just-in-time Compilation based Query Processing on GPUs	2021	VLDB	6.0805143e-05
6,762	SquirrelJoin: Network-Aware Distributed Join Processing with Lazy Partitioning	2017	VLDB	5.7814194e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	2,140	Revisiting Co-Processing for Hash Joins on the Coupled CPU-GPU Architecture	2013	VLDB
2	360	Design and Evaluation of Main Memory Hash Join Algorithms for Multi-core CPUs	2011	SIGMOD
3	209	Sort vs. Hash Revisited: Fast Join Implementation on Modern Multi-Core CPUs	2009	VLDB
4	2,566	Pump Up the Volume: Processing Large Data on GPUs with Fast Interconnects	2020	SIGMOD
5	7,591	Efficiently Processing Joins and Grouped Aggregations on GPUs	2025	SIGMOD
6	4,639	MQJoin: Efficient Shared Execution of Main-Memory Joins	2016	VLDB
7	631	Relational Joins on Graphics Processors	2008	SIGMOD
8	4,535	Triton Join: Efficiently Scaling to a Large Join State on GPUs with Fast Interconnects	2022	SIGMOD
9	3,134	Efficient Join Algorithms For Large Database Tables in a Multi-GPU Environment	2021	VLDB
10	9,333	Efficiently Joining Large Relations on Multi-GPU Systems	2025	VLDB