Locally-Adaptive Quantization for Streaming Vector Search

TL;DR

This paper evaluates Locally-Adaptive Vector Quantization (LVQ) and its new variants in streaming similarity search, demonstrating significant performance improvements and robustness to data distribution shifts, and introduces an open-source library for high-performance vector search.

Contribution

The paper introduces Turbo LVQ and multi-means LVQ, enhancing LVQ's performance in streaming similarity search and providing the first comprehensive evaluation of LVQ in evolving data scenarios.

Findings

LVQ and variants outperform competitors by up to 9.4x in static data.

LVQ variants improve search speed by up to 28%.

Robust performance under data distribution shifts.

Abstract

Retrieving the most similar vector embeddings to a given query among a massive collection of vectors has long been a key component of countless real-world applications. The recently introduced Retrieval-Augmented Generation is one of the most prominent examples. For many of these applications, the database evolves over time by inserting new data and removing outdated data. In these cases, the retrieval problem is known as streaming similarity search. While Locally-Adaptive Vector Quantization (LVQ), a highly efficient vector compression method, yields state-of-the-art search performance for non-evolving databases, its usefulness in the streaming setting has not been yet established. In this work, we study LVQ in streaming similarity search. In support of our evaluation, we introduce two improvements of LVQ: Turbo LVQ and multi-means LVQ that boost its search performance by up to 28% and 27%, respectively. Our studies show that LVQ and its new variants enable blazing fast vector search, outperforming its closest competitor by up to 9.4x for identically distributed data and by up to 8.8x under the challenging scenario of data distribution shifts (i.e., where the statistical distribution of the data changes over time). We release our contributions as part of Scalable Vector Search, an open-source library for high-performance similarity search.