GPU-Accelerated ANNS: Quantized for Speed, Built for Change
Summary: Jasper is a GPU-native, fully updatable Vamana ANNS index, combining lock-free batch-parallel insertions with GPU-efficient RaBitQ quantization (up to 8× smaller). Its latency-hiding search kernel delivers up to 1.84× CAGRA throughput and 7× faster construction. (summarized by gpt-5.6-luna on Aug 28 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Hunter McCoy (Northeastern University)
- 2. Zikun Wang (Northeastern University)
- 3. Prashant Pandey (Northeastern University)
BibTeX Citation
@article{mccoy_vldb26,
title = {{GPU-Accelerated ANNS: Quantized for Speed, Built for Change}},
author = {McCoy, Hunter and Wang, Zikun and Pandey, Prashant},
journal = {PVLDB},
series = {{VLDB} '26},
volume = {19},
number = {11},
pages = {3413--3426},
doi = {10.14778/3836663.3836698},
url = {https://doi.org/10.14778/3836663.3836698},
year = {2026}
}Incoming Citations (Sorted by Pagerank)
Showing 1 of 1 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 10,873 | GPU-Native Approximate Nearest Neighbor Search with IVF-RaBitQ: Fast Index Build and Search | 2026 | VLDB | 4.9793485e-05 |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 2 of 2 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 804 | RaBitQ: Quantizing High-Dimensional Vectors with a Theoretical Error Bound for Approximate Nearest Neighbor Search | 2024 | SIGMOD | 0.00013832333 |
| 5,411 | GTS: GPU-based Tree Index for Fast Similarity Search | 2024 | SIGMOD | 6.1408042e-05 |
Previous
Page 1 / 1
Next