VecFlow-Chamfer: A GPU-based Data Management System for High-Performance Multi-Vector Search on Superchips
Summary: GPU-native multi-vector search for token-level retrieval on Superchips: compression-free MaxIVF-CAGRA + ChamferCore + GraceStore deliver high-recall candidate gen, millisecond Chamfer scoring, and on-demand full-precision vector access. Order-of-magnitude lower latency vs PLAID/MUVERA while improving recall. (summarized by gpt-5.4-mini on Apr 11 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Chenghao Mo (University of Illinois Urbana-Champaign)
- 2. Ben Karsin (NVIDIA)
- 3. Philip Adams (Microsoft)
- 4. Minjia Zhang (University of Illinois Urbana-Champaign)
BibTeX Citation
@inproceedings{mo_sigmod26,
title = {{VecFlow-Chamfer: A GPU-based Data Management System for High-Performance Multi-Vector Search on Superchips}},
author = {Mo, Chenghao and Karsin, Ben and Adams, Philip and Zhang, Minjia},
series = {{SIGMOD} '26},
booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
publisher = {Association for Computing Machinery},
doi = {10.1145/3786706},
url = {https://dl.acm.org/doi/10.1145/3786706},
year = {2026}
}Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 0 of 0 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next