Skip to main content
Ctrl+K
cuvs - Home cuvs - Home

cuvs

  • GitHub
  • Twitter
Home
cuvs
cucimcudf-javacudfcugraphcumlcuprojcuspatialcuvscuxfilterdask-cudadask-cudfkvikiolibcudflibcumllibcuprojlibcuspatiallibkvikiolibrmmlibucxxraftrapids-cmakerapidsmpfrmm
legacy (25.04)
nightly (25.08)stable (25.06)legacy (25.04)
cuvs - Home cuvs - Home

cuvs

  • GitHub
  • Twitter

Table of Contents

  • Installation
  • Getting Started
    • Primer on vector search indexes
    • Vector search indexes vs vector databases
    • Automated tuning Guide
    • Comparing performance of vector indexes
    • Nearest Neighbor Indexes
      • Brute-force
      • CAGRA
      • IVF-Flat
      • IVF-PQ
      • Vamana
    • cuVS API Basics
    • Interoperability
    • Working with ANN Indexes
      • Working with ANN Indexes in C
      • Working with ANN Indexes in C++
      • Working with ANN Indexes in Python
      • Working with ANN Indexes in Rust
    • Filtering vector indexes
  • Integrations
    • FAISS
    • Milvus
    • Lucene
    • Kinetica
  • cuVS Bench
  • API Reference
    • C API Documentation
      • Core Routines
      • Nearest Neighbors
        • Bruteforce
        • IVF-Flat
        • IVF-PQ
        • CAGRA
        • HNSW
    • C++ API Documentation
      • Cluster
        • Cluster
        • Cluster
      • Distance
      • Nearest Neighbors
        • Bruteforce
        • CAGRA
        • Dynamic Batching
        • Filtering
        • HNSW
        • IVF-Flat
        • IVF-PQ
        • NN-Descent
        • Refinement
        • Distributed ANN
        • Vamana
      • Preprocessing
        • Quantize
      • Selection
      • Stats
    • Python API Documentation
      • Distance
      • Nearest Neighbors
        • Brute Force KNN
        • CAGRA
        • HNSW
        • IVF-Flat
        • IVF-PQ
        • NN-Descent
      • Preprocessing
    • Rust API Documentation
  • Contributing
  • Getting Started
  • Nearest Neighbor Indexes

Nearest Neighbor Indexes#

Contents:

  • Brute-force
    • Filtering considerations
    • Configuration parameters
      • Build parameters
      • Search Parameters
    • Tuning Considerations
    • Memory footprint
      • Index footprint
  • CAGRA
    • Interoperability with HNSW
    • Filtering considerations
    • Configuration parameters
      • Build parameters
      • Search parameters
    • Tuning Considerations
    • Memory footprint
      • Dataset on device (graph on host):
      • Dataset on host (graph on host):
      • Build peak memory usage:
      • Build with out-of-core IVF-PQ peak memory usage:
  • IVF-Flat
    • Filtering considerations
    • Configuration parameters
      • Build parameters
      • Search parameters
    • Tuning Considerations
    • Memory footprint
      • Index (device memory):
      • Peak device memory usage for index build:
  • IVF-PQ
    • Configuration parameters
      • Build parameters
      • Search parameters
    • Tuning Considerations
    • Memory footprint
      • Index (device memory):
      • Index (host memory):
      • Search peak memory usage (device);
      • Build peak memory usage (device):
  • Vamana
    • Interoperability with CPU DiskANN
    • Configuration parameters
      • Build parameters
    • Tuning Considerations
    • Memory footprint
      • Device memory usage

Indices and tables#

  • Index

  • Module Index

  • Search Page

previous

Comparing performance of vector indexes

next

Brute-force

On this page
  • Nearest Neighbor Indexes
  • Indices and tables
NVIDIA NVIDIA
Privacy Policy | Manage My Privacy | Do Not Sell or Share My Data | Terms of Service | Accessibility | Corporate Policies | Product Security | Contact

Copyright © 2024, NVIDIA Corporation.