Aisaq: All-in-storage ANNS With Product Quantization For Dram-free Information Retrieval

Abstract

Graph-based approximate nearest neighbor search (ANNS) algorithms work effectively against large-scale vector retrieval. Among such methods, DiskANN achieves good recall-speed tradeoffs using both DRAM and storage. DiskANN adopts product quantization (PQ) to reduce memory usage, which is still proportional to the scale of datasets. In this paper, we propose All-in-Storage ANNS with Product Quantization (AiSAQ), which offloads compressed vectors to the SSD index. Our method achieves \(\sim\)10 MB memory usage in query search with billion-scale datasets without critical latency degradation. AiSAQ also reduces the index load time for query search preparation, which enables fast switch between muitiple billion-scale indices.This method can be applied to retrievers of retrieval-augmented generation (RAG) and be scaled out with multiple-server systems for emerging datasets. Our DiskANN-based implementation is available on GitHub.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keytatsuno2024aisaq

Related papers

• Diskann++: Efficient Page-based Search Over Isomorphic Mapped Graph Index Using Query-sensitivity Entry Vertex (2023)0.00
• SPANN: Highly-efficient Billion-scale Approximate Nearest Neighbor Search (2021)0.00
• Routing-guided Learned Product Quantization For Graph-based Approximate Nearest Neighbor Search (2023)4.52
• DGAI: Decoupled On-disk Graph-based ANN Index For Efficient Updates And Queries (2025)0.00
• Online Product Quantization (2017)10.61
• SAQ: Pushing The Limits Of Vector Quantization Through Code Adjustment And Dimension Segmentation (2025)0.00
• Ood-diskann: Efficient And Scalable Graph ANNS For Out-of-distribution Queries (2022)0.00
• Rabitq: Quantizing High-dimensional Vectors With A Theoretical Error Bound For Approximate Nearest Neighbor Search (2024)12.54