Scalable Asynchronous Gradient Descent Optimization for Out-of-Core Models

Summary: Scales asynchronous SGD for out-of-core models via vertical offline partitioning of the model and online updates. A preemptive push-based sharing mechanism minimizes disk I/O, delivering improved convergence over HOGWILD! and enabling scalability to massive models. (summarized by gpt-5-nano on Feb 09 2026)

Paper ID: 11780
Venue: VLDB
Year: 2017
Pagerank: 6.5024714e-05
Overall Rank: 4,799 | 67.08%
DOI: 10.14778/3115404.3115405

Incoming Non-self Citations Over Time

Authors

1. Chengjie Qin (GraphSQL)
2. Martin Torres (University of California Merced)
3. Florin Rusu (University of California Merced)

BibTeX Citation

@article{qin_vldb17,
        title = {{Scalable Asynchronous Gradient Descent Optimization for Out-of-Core Models}},
        author = {Qin, Chengjie and Torres, Martin and Rusu, Florin},
        journal = {PVLDB},
        series = {{VLDB} '17},
        volume = {10},
        number = {10},
        pages = {986--997},
        doi = {10.14778/3115404.3115405},
        url = {https://doi.org/10.14778/3115404.3115405},
        year = {2017}
}

Incoming Citations (Sorted by Pagerank)

Showing 3 of 3 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
2,693	FlexPS: Flexible Parallelism Control in Parameter Server Architecture	2018	VLDB	8.247004e-05
3,782	CROSSBOW: Scaling Deep Learning with Small Batch Sizes on Multi-GPU Servers	2019	VLDB	7.1298979e-05
6,485	Tuple-oriented Compression for Large-scale Mini-batch Stochastic Gradient Descent	2019	SIGMOD	5.8657457e-05

Outgoing Citations (Sorted by Pagerank)

Showing 12 of 12 cited papers.

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

Rank	Cited Paper	Year	Venue	Pagerank
106	The MADlib Analytics Library or MAD Skills, the SQL	2012	VLDB	0.00033539462
518	Towards a Unified Architecture for in-RDBMS Analytics	2012	SIGMOD	0.00017167492
536	Learning Linear Regression Models over Factorized Joins	2016	SIGMOD	0.0001693369
640	Materialization Optimizations for Feature Selection Workloads	2014	SIGMOD	0.00015409494
715	Learning Generalized Linear Models Over Normalized Data	2015	SIGMOD	0.00014655327
764	To Join or Not to Join? Thinking Twice about Joins before Feature Selection	2016	SIGMOD	0.00014226652
993	Simulation of Database-Valued Markov Chains Using SimSQL	2013	SIGMOD	0.00012789598
1,294	Augmented Sketch: Faster and More Accurate Stream Processing	2016	SIGMOD	0.00011291308
1,632	Data Morphing: An Adaptive, Cache-Conscious Storage Technique	2003	VLDB	0.0001016847
4,214	The Missing Piece in Complex Analytics: Low Latency, Scalable Model Management and Serving with Velox	2015	CIDR	6.8276099e-05
4,897	GLADE: Big Data Analytics Made Easy	2012	SIGMOD	6.4546134e-05
7,059	A Comparison of Knives for Bread Slicing	2013	VLDB	5.7146962e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	2,640	Scalable and Efficient Full-Graph GNN Training for Large Graphs	2023	SIGMOD
2	7,000	A Cost-based Optimizer for Gradient Descent Optimization	2017	SIGMOD
3	9,290	Hyper-Tune: Towards Efficient Hyper-parameter Tuning at Scale	2022	VLDB
4	9,371	Towards an Optimized GROUP BY Abstraction for Large-Scale Machine Learning	2021	VLDB
5	6,339	In-Database Machine Learning with CorgiPile: Stochastic Gradient Descent without Full Data Shuffle	2022	SIGMOD
6	4,697	Multi-Dimensional Balanced Graph Partitioning via Projected Gradient Descent	2019	VLDB
7	1,079	Hybrid Parallelization Strategies for Large-Scale Machine Learning in SystemML	2014	VLDB
8	9,729	Scalable Graph Convolutional Network Training on Distributed-Memory Systems	2023	VLDB
9	2,162	Heterogeneity-aware Distributed Parameter Servers	2017	SIGMOD
10	4,956	Heterogeneity-Aware Distributed Machine Learning Training via Partial Reduce	2021	SIGMOD