Pg_vectorsearch: An index that is faster than pgvector's IVFFlat and HNSW

Pg_vectorsearch is a PostgreSQL extension featuring a new PRISM index for approximate nearest neighbor search that outperforms pgvector's existing IVFFlat and HNSW indexes in both query speed and index building efficiency. PRISM uses hierarchical partitioning, RaBitQ quantization with error bounds, and SIMD-optimized distance computation to achieve faster performance even with reduced memory requirements compared to HNSW on large-scale datasets. The extension is currently in alpha status and is compatible with existing pgvector columns.