Similarity search in the blink of an eye with compressed indices
Summary: LVQ—per-vector scaling plus scalar quantization—for graph-based indices, cuts memory and effective bandwidth with negligible accuracy loss to accelerate similarity computations. With a new high-performance graph engine, LVQ is SOTA on billion-scale search: up to 20.7× throughput with ~3× lower memory, and 5.8× with 1.4× memory savings. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Cecilia Aguerrebere (Intel)
- 2. Ishwar Singh Bhati (Intel)
- 3. Mark Hildebrand (Intel)
- 4. Mariano Tepper (Intel)
- 5. Theodore Willke (Intel)
BibTeX Citation
@article{aguerrebere_vldb23,
title = {{Similarity search in the blink of an eye with compressed indices}},
author = {Aguerrebere, Cecilia and Bhati, Ishwar Singh and Hildebrand, Mark and Tepper, Mariano and Willke, Theodore},
journal = {PVLDB},
series = {{VLDB} '23},
volume = {16},
number = {11},
pages = {3433--3446},
doi = {10.14778/3611479.3611537},
url = {https://doi.org/10.14778/3611479.3611537},
year = {2023}
}
Incoming Citations (Sorted by Pagerank)
Showing 23 of 23 citing papers.
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 4 of 4 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 93 | Fast Approximate Nearest Neighbor Search With The Navigating Spreading-out Graph | 2019 | VLDB | 0.00034701237 |
| 398 | A Comprehensive Survey and Experimental Comparison of Graph-Based Approximate Nearest Neighbor Search | 2021 | VLDB | 0.00019194947 |
| 805 | Cache locality is not enough: High-Performance Nearest Neighbor Search with Product Quantization Fast Scan | 2016 | VLDB | 0.00013891999 |
| 2,572 | DeltaPQ: Lossless Product Quantization Code Compression for High Dimensional Similarity Search | 2020 | VLDB | 8.4027322e-05 |
Previous
Page 1 / 1
Next