VecFlow-Chamfer: A GPU-based Data Management System for High-Performance Multi-Vector Search on Superchips
Proceedings of the ACM on Management of Data · 02.04.2026 · 4 (1) · 1-26
DOI: 10.1145/3786706 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the ACM on Management of Data, online erschienen am 07.04.2026. Verfasst von Chenghao Mo und 3 weiteren Autor:innen. Die Arbeit nennt 43 Literaturangaben.
Abstract (Originalsprache)
Many emerging AI applications demand retrieval systems that go beyond document-level relevance and capture token-level semantics. Multi-vector search addresses this need through fine-grained semantic matching between token-level embeddings of queries and documents. However, it introduces significant system challenges, including compute-intensive set-to-set scoring, complex candidate filtering, and high memory overhead from storing dense token-level embeddings. Prior systems mitigate these challenges through GPU-based similarity calculation and indexing structures (e.g., IVFPQ-GPU) but often at the cost of reduced retrieval accuracy and suffer from low GPU utilization. We present \name, a GPU-based vector data management system that enables low-latency and high-recall multi-vector search on modern Superchip architectures. \name achieves this through a combination of three novel optimizations: (1) \maxivf, a GPU-native, compression-free index tailored for multi-vector search with fine-grained anchor vectors enabling scalable index construction and low-latency, high-accuracy candidate generation via CAGRA-based GPU routing; (2) \chamferkernel, a highly-optimized GPU kernel that enables single-digit millisecond Chamfer scoring over tens of thousands of candidate documents; and (3) \zerocomp, a tiered vector storage layer that supports on-demand, low-latency access to full-precision embeddings across Grace-Hopper NVLink-C2C interconnects. Together, these techniques enable \name to perform multi-vector search over hundreds of millions of document token embeddings with unprecedented high recall and low latency. Compared to state-of-the-art systems like PLAID and MUVERA, \name achieves an order-of-magnitude lower latency while significantly improving recall.
Zitieren
APA
Mo, C., Karsin, B., Adams, P., & Zhang, M. (2026). VecFlow-Chamfer: A GPU-based Data Management System for High-Performance Multi-Vector Search on Superchips. Proceedings of the ACM on Management of Data, 4(1), 1-26. https://doi.org/10.1145/3786706
Vancouver
Mo C, Karsin B, Adams P, Zhang M. VecFlow-Chamfer: A GPU-based Data Management System for High-Performance Multi-Vector Search on Superchips. Proceedings of the ACM on Management of Data. 2026;4(1):1-26. doi:10.1145/3786706
BibTeX
@article{Mo2026,
author = {Mo, Chenghao and Karsin, Ben and Adams, Philip and Zhang, Minjia},
title = {VecFlow-Chamfer: A GPU-based Data Management System for High-Performance Multi-Vector Search on Superchips},
journal = {Proceedings of the ACM on Management of Data},
year = {2026},
volume = {4},
number = {1},
pages = {1--26},
doi = {10.1145/3786706}
}
EUVIMED