GateANN: I/O-Efficient Filtered Vector Search on SSDs

2026-03-22

AI Summary

This work addresses the inefficiency of existing SSD-based approximate nearest neighbor (ANN) search systems when handling filtered queries, which often suffer from excessive invalid I/O due to post-filtering or require frequent index rebuilds to accommodate filtering conditions. To overcome these limitations, the authors propose GateANN, a novel approach that introduces a graph tunneling mechanism. This mechanism dynamically skips non-matching nodes in memory by jointly evaluating approximate distances and filter predicates during traversal, without modifying the underlying graph index. By decoupling graph traversal from vector retrieval, GateANN avoids issuing SSD accesses to irrelevant nodes. Experimental results demonstrate that GateANN reduces SSD read volume by up to 10× and achieves up to 7.6× higher throughput compared to state-of-the-art methods.

Technology Category

• Search and Optimization: Distributed Search
• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community

Application Category

• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Social Networks and Social Media: Social mining and social search on the Web

Abstract

We present GateANN, an I/O-efficient SSD-based graph ANNS system that supports filtered vector search on an unmodified graph index. Existing SSD-based systems either waste I/O by post-filtering, or require expensive filter-aware index rebuilds. GateANN avoids both by decoupling graph traversal from vector retrieval. Our key insight is that traversing a node requires only its neighbor list and an approximate distance, neither of which needs the full-precision vector on SSD. Based on this, GateANN introduces graph tunneling. It checks each node's filter predicate in memory before issuing I/O and routes through non-matching nodes entirely in memory, preserving graph connectivity without any SSD read for non-matching nodes. Our experimental results show that it reduces SSD reads by up to 10x and improves throughput by up to 7.6x.

Problem

• filtered vector search
• SSD-based ANNS
• I/O efficiency
• graph index
• vector retrieval

Innovation

• filtered vector search
• graph ANNS
• I/O efficiency
• SSD-based retrieval
• graph tunneling