Starling: An I/O-Efficient Disk-Resident Graph Index Framework for High-Dimensional Vector Similarity Search on Data Segment
Summary: Starling: a disk-resident, I/O-efficient HVSS framework for segment-wide vector search. Hybrid layout (in-memory navigation graph + reordered disk graph) and a block I/O strategy reduce disk traffic, enabling 33M 128-D vectors with >0.9 AP and ~1 ms latency, ~44x throughput vs prior methods. (summarized by gpt-5-nano on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Mengzhao Wang (Zhejiang University)
- 2. Weizhi Xu (Zilliz)
- 3. Xiaomeng Yi (Zhejiang Lab)
- 4. Songlin Wu (Tongji University)
- 5. Zhangyang Peng (Hangzhou Dianzi University)
- 6. Xiangyu Ke (Zhejiang University)
- 7. Yunjun Gao (Zhejiang University)
- 8. Xiaoliang Xu (Hangzhou Dianzi University)
- 9. Rentong Guo (Zilliz)
- 10. Charles Xie (Zilliz)
BibTeX Citation
@inproceedings{wang_sigmod24,
title = {{Starling: An I/O-Efficient Disk-Resident Graph Index Framework for High-Dimensional Vector Similarity Search on Data Segment}},
author = {Wang, Mengzhao and Xu, Weizhi and Yi, Xiaomeng and Wu, Songlin and Peng, Zhangyang and Ke, Xiangyu and Gao, Yunjun and Xu, Xiaoliang and Guo, Rentong and Xie, Charles},
series = {{SIGMOD} '24},
booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
publisher = {Association for Computing Machinery},
doi = {10.1145/3639269},
url = {https://dl.acm.org/doi/10.1145/3639269},
year = {2024}
}
Incoming Citations (Sorted by Pagerank)
Showing 35 of 35 citing papers.
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 18 of 18 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next