cedar: Optimized and Unified Machine Learning Input Data Pipelines

Summary: Cedar provides a unified, composable framework for ML input pipelines across arbitrary frameworks, with an extensible optimizer that combines execution optimizations. It orchestrates local/distributed resources and delivers 1.87–10.65× speedups over existing systems. (summarized by gpt-5.6-luna on Jul 24 2026)

Paper ID: 14280
Venue: VLDB
Year: 2025
Pagerank: 5.093636e-05
Overall Rank: 10,999 | 24.54%
DOI: 10.14778/3705829.3705861

Incoming Non-self Citations Over Time

No non-self incoming citations found for this paper in this database.

Authors

1. Mark Zhao (Stanford University)
2. Emanuel Adamiak (Stanford University)
3. Christos Kozyrakis (Stanford University)

BibTeX Citation

@article{zhao_vldb25,
        title = {{cedar: Optimized and Unified Machine Learning Input Data Pipelines}},
        author = {Zhao, Mark and Adamiak, Emanuel and Kozyrakis, Christos},
        journal = {PVLDB},
        series = {{VLDB} '25},
        volume = {18},
        number = {2},
        pages = {488--502},
        doi = {10.14778/3705829.3705861},
        url = {https://doi.org/10.14778/3705829.3705861},
        year = {2025}
}

Incoming Citations (Sorted by Pagerank)

Showing 0 of 0 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank

Outgoing Citations (Sorted by Pagerank)

Showing 13 of 13 cited papers.

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

Rank	Cited Paper	Year	Venue	Pagerank
24	Spark SQL: Relational Data Processing in Spark	2015	SIGMOD	0.00054865648
66	The Snowflake Elastic Data Warehouse	2016	SIGMOD	0.00038561587
292	LINQ: Reconciling Objects, Relations and XML in the .NET Framework	2006	SIGMOD	0.00022259549
361	The Dataflow Model: A Practical Approach to Balancing Correctness, Latency, and Cost in Massive-Scale, Unbounded, Out-of-Order Data Processing	2015	VLDB	0.00020138717
520	Delta Lake: High-Performance ACID Table Storage over Cloud Object Stores	2020	VLDB	0.00017136828
1,446	Analyzing and Mitigating Data Stalls in DNN Training	2021	VLDB	0.0001076818
2,018	tf.data: A Machine Learning Data Processing Framework	2021	VLDB	9.3001937e-05
2,164	Opening the Black Boxes in Data Flow Optimization	2012	VLDB	9.0521951e-05
2,239	An Intermediate Representation for Optimizing Machine Learning Pipelines	2019	VLDB	8.8875753e-05
3,541	Where Is My Training Bottleneck? Hidden Trade-Offs in Deep Learning Preprocessing Pipelines	2022	SIGMOD	7.3280673e-05
3,764	FastFlow: Accelerating Deep Learning Model Training with Smart Offloading of Input Data Pipeline	2023	VLDB	7.1446931e-05
5,282	GoldMiner: Elastic Scaling of Training Data Pre-Processing Pipelines for Deep Learning	2023	SIGMOD	6.2840712e-05
8,321	FusionFlow: Accelerating Data Preprocessing for Machine Learning with CPU-GPU Cooperation	2024	VLDB	5.4544352e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	6,309	Materialization and Reuse Optimizations for Production Data Science Pipelines	2022	SIGMOD
2	1,157	Cerebro: A Data System for Optimized Deep Learning Model Selection	2020	VLDB
3	8,949	Modyn: Data-Centric Machine Learning Pipeline Orchestration	2025	SIGMOD
4	3,541	Where Is My Training Bottleneck? Hidden Trade-Offs in Deep Learning Preprocessing Pipelines	2022	SIGMOD
5	4,592	Data Platform for Machine Learning	2019	SIGMOD
6	4,240	LIMA: Fine-grained Lineage Tracing and Reuse in Machine Learning Systems	2021	SIGMOD
7	8,101	SDP_PIPE: A Semi-Decentralized Framework for Heterogeneity-aware Pipeline-parallel Training	2023	VLDB
8	7,160	DAPHNE: An Open and Extensible System Infrastructure for Integrated Data Analysis Pipelines	2022	CIDR
9	10,684	Sequoia: An Accessible and Extensible Framework for Privacy-Preserving Machine Learning over Distributed Data	2025	SIGMOD
10	2,018	tf.data: A Machine Learning Data Processing Framework	2021	VLDB