# Disk Based

> Browse 9 items tagged with Disk Based.

## Items

- [HNSW-IF](https://vectordb.ever.works/items/hnsw-if) — Hybrid billion-scale vector search method combining HNSW with inverted file indexes, enabling cost-efficient search by keeping centroids in memory while storing vectors on disk.
- [BatANN: Passing the Baton: High Throughput Distributed Disk-Based Vector Search](https://vectordb.ever.works/items/batann-passing-the-baton-high-throughput-distributed-disk-based-vector-search) — BatANN system by Dang et al. for high-throughput distributed disk-based vector search. Supports scalable ANN in distributed environments.
- [Passing the Baton: High Throughput Distributed Disk-Based Vector Search with BatANN](https://vectordb.ever.works/items/passing-the-baton-high-throughput-distributed-disk-based-vector-search-with-batann) — A distributed, disk-based vector search system designed for high-throughput approximate nearest neighbor queries at scale. BatANN provides an architecture and methods applicable to large-scale vector databases that need efficient storage beyond memory, enabling cost-effective approximate nearest neighbor search for high-dimensional embeddings.
- [Starling — I/O-Efficient Disk-Resident Graph Index Framework](https://vectordb.ever.works/items/starling-io-efficient-disk-resident-graph-index-framework) — SIGMOD 2024 paper introducing Starling, an I/O-efficient disk-resident graph index framework for high-dimensional vector similarity search on data segments, optimizing disk access patterns for billion-scale datasets.
- [OrchANN](https://vectordb.ever.works/items/orchann-io-framework) — A unified I/O orchestration framework for skewed out-of-core vector search that addresses the challenge of billion-scale ANN search when the dataset exceeds available memory. OrchANN optimizes I/O operations for graph-based indexes stored on disk.
- [MCGI](https://vectordb.ever.works/items/mcgi) — Manifold-Consistent Graph Indexing for billion-scale disk-resident vector search. Leverages Local Intrinsic Dimensionality to achieve 5.8x throughput improvement over DiskANN on high-dimensional datasets.
- [OdinANN](https://vectordb.ever.works/items/odinann) — Billion-scale graph-based ANNS index with direct insertion capabilities. Achieves <1ms search latency with >10x less memory than in-memory indexes through GC-free design and update combining.
- [PageANN](https://vectordb.ever.works/items/pageann) — Disk-based approximate nearest neighbor search framework with page-aligned graph structure. Achieves 1.85x-10.83x higher throughput than state-of-the-art methods through optimized SSD utilization.
- [PipeANN](https://vectordb.ever.works/items/pipeann) — Low-latency, billion-scale updatable graph-based vector store on SSD. Achieves <1ms search latency with 10x less memory than in-memory indexes through alignment of best-first search with SSD characteristics.

---

_Canonical page: https://vectordb.ever.works/tags/disk-based_