IVF-PQ
An approximate nearest neighbor (ANN) indexing method that combines an inverted index with product quantization. By partitioning the vector space into multiple clusters (inverted files) and applying product quantization to sub-vectors for compression, it achieves fast approximate vector retrieval while significantly reducing memory usage, making it suitable for large-scale vector scenarios.
Core Facts
Timeline (last 90 days)
The core trade-off of ANN indexes (such as HNSW, IVF-PQ) lies in the triangular relationship among recall, memory and build time.
All Facts (2)
Approximate nearest neighbor (ANN) algorithms such as HNSW and IVF achieve order-of-magnitude speed improvements by sacrificing a small amount of precision.
65%UnverifiedThe core trade-off of ANN indexes (such as HNSW, IVF-PQ) lies in the triangular relationship among recall, memory and build time.
50%