Fast Vector Search in PostgreSQL: A Decoupled Approach

Summary: PostgreSQL-V decouples vector indexes from PostgreSQL's page-oriented core to directly leverage native high-performance vector index libraries, avoiding legacy overhead of prior integrated designs. Employs a lightweight consistency mechanism and matches specialized DB speed, up to 8.9x vs pgvector. (summarized by gpt-5-mini on Feb 09 2026)

Paper ID: hc577101682e27488
Venue: CIDR
Year: 2026
Pagerank: 5.4169862e-05
Overall Rank: 7,957 | 46.53%
DOI: -
PDF: Download (CC BY 4.0)

Incoming Non-self Citations Over Time

Authors

1. Jiayi Liu (Purdue University)
2. Yunan Zhang (Purdue University)
3. Chenzhe Jin (Purdue University)
4. Aditya Gupta (Purdue University)
5. Shige Liu (Purdue University)
6. Jianguo Wang (Purdue University)

BibTeX Citation

@inproceedings{liu_cidr26,
        address = {Amsterdam, Netherlands},
        series = {{CIDR} '26},
        title = {{Fast Vector Search in PostgreSQL: A Decoupled Approach}},
        booktitle = {Proceedings of the {Conference} on {Innovative} {Data} {Systems} {Research}},
        author = {Liu, Jiayi and Zhang, Yunan and Jin, Chenzhe and Gupta, Aditya and Liu, Shige and Wang, Jianguo},
        year = {2026}
}

Incoming Citations (Sorted by Pagerank)

Showing 5 of 5 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
10,421	An In-Depth Study of Filter-Agnostic Vector Search on a PostgreSQL Database System: [Experiments & Analysis]	2026	SIGMOD	4.9769913e-05
10,527	VecBench: A Controllable Benchmark for Filtered Vector Search: [Experiments & Analysis]	2026	SIGMOD	4.9769913e-05
10,648	Efficient Vector Index Merging in Vector Databases	2026	SIGMOD	4.9769913e-05
10,689	Reducing Tail Latency in Storage-Disaggregated Database Systems	2026	SIGMOD	4.9769913e-05
10,957	Nova: A Multi-Purpose Vector Engine for Low-Latency, Multi-Tenant, and Cross-Table Hybrid Retrieval	2026	VLDB	4.9769913e-05

Outgoing Citations (Sorted by Pagerank)

Showing 9 of 9 cited papers.

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

Rank	Cited Paper	Year	Venue	Pagerank
63	Amazon Aurora: Design Considerations for High Throughput Cloud-Native Relational Databases	2017	SIGMOD	0.00038521221
189	Milvus: A Purpose-Built Vector Data Management System	2021	SIGMOD	0.0002585319
339	AnalyticDB-V: A Hybrid Analytical Engine Towards Query Fusion for Structured and Unstructured Data	2020	VLDB	0.00020561716
543	Socrates: The New SQL Server in the Cloud	2019	SIGMOD	0.00016624903
909	PASE: PostgreSQL Ultra-High-Dimensional Approximate Nearest Neighbor Search Extension	2020	SIGMOD	0.0001312576
1,979	SingleStore-V: An Integrated Vector Database System in SingleStore	2024	VLDB	9.271804e-05
7,079	Understanding the Performance Implications of the Design Principles in Storage-Disaggregated Databases	2024	SIGMOD	5.603281e-05
7,239	CaaS-LSM: Compaction-as-a-Service for LSM-based Key-Value Stores in Storage Disaggregated Infrastructure	2024	SIGMOD	5.5761767e-05
10,170	Vector Databases: What’s Really New and What’s Next? (VLDB 2024 Panel)	2024	VLDB	5.0658661e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	11,037	Advances of Query Processing in Vector Databases	2026	VLDB
2	10,527	VecBench: A Controllable Benchmark for Filtered Vector Search: [Experiments & Analysis]	2026	SIGMOD
3	9,039	Cost-Effective, Low Latency Vector Search with Azure Cosmos DB	2025	VLDB
4	10,919	TEngineDB-V: An OLAP-Native Vector Search System for Large-k Workloads at Tencent	2026	VLDB
5	10,648	Efficient Vector Index Merging in Vector Databases	2026	SIGMOD
6	8,413	GaussDB-Vector: A Large-Scale Persistent Real-Time Vector Database for LLM Applications	2025	VLDB
7	10,938	SQL-Native Vector Search at Billion Scale in Presto	2026	VLDB
8	1,979	SingleStore-V: An Integrated Vector Database System in SingleStore	2024	VLDB
9	8,654	Turbocharging Vector Databases using Modern SSDs	2025	VLDB
10	10,421	An In-Depth Study of Filter-Agnostic Vector Search on a PostgreSQL Database System: [Experiments & Analysis]	2026	SIGMOD