GPS: Revisiting the Data Layout for Disk-based High-Dimensional Vector Search
Summary: GPS revisits SSD-based vector-index layout by prioritizing frequently accessed proximity-graph adjacency lists over vectors. Its graph-aware cache and neighbor-co-located disk blocks improve locality, yielding 52% higher throughput and 32% lower latency than PipeANN. (summarized by gpt-5.6-luna on Jul 26 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Peiqi Yin (Chinese University of Hong Kong)
- 2. Xiao Yan (Wuhan University)
- 3. Qihui Zhou (Chinese University of Hong Kong)
- 4. Hui Li (Chinese University of Hong Kong)
- 5. Xiaolu Li (Huazhong University of Science and Technology)
- 6. Meiling Wang (Huawei)
- 7. Lin Zhang (Huawei)
- 8. Xin Yao (Huawei)
- 9. James Cheng (Chinese University of Hong Kong)
BibTeX Citation
@inproceedings{yin_sigmod26,
title = {{GPS: Revisiting the Data Layout for Disk-based High-Dimensional Vector Search}},
author = {Yin, Peiqi and Yan, Xiao and Zhou, Qihui and Li, Hui and Li, Xiaolu and Wang, Meiling and Zhang, Lin and Yao, Xin and Cheng, James},
series = {{SIGMOD} '26},
booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
publisher = {Association for Computing Machinery},
doi = {10.1145/3802069},
url = {https://dl.acm.org/doi/10.1145/3802069},
year = {2026}
}Incoming Citations (Sorted by Pagerank)
Showing 1 of 1 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 9,889 | SAQ: Pushing the Limits of Vector Quantization through Code Adjustment and Dimension Segmentation | 2026 | SIGMOD | 5.1997534e-05 |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 27 of 27 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next