Topics
Browse everything VectorOpsReport publishes by category and tag: every topic covered here, with the most recent guides listed under each one.
Tags
- #hnsw 10
- #vector-database 8
- #recall 6
- #vector-search 5
- #pgvector 4
- #ann 3
- #ann-search 3
- #faiss 3
- #quantization 3
- #comparison 2
- #embeddings 2
- #mlops 2
- #rag 2
- #benchmarking 1
- #benchmarks 1
- #bm25 1
- #capacity-planning 1
- #cosine-similarity 1
- #dot-product 1
- #ef-search 1
- #filtering 1
- #hybrid-search 1
- #index-selection 1
- #index-sizing 1
- #ivf 1
- #m-parameter 1
- #memory-sizing 1
- #milvus 1
- #observability 1
- #pinecone 1
- #postgresql 1
- #product-quantization 1
- #qdrant 1
- #reciprocal-rank-fusion 1
- #reranking 1
- #troubleshooting 1
- #weaviate 1
Categories
Comparison 2 posts
- pgvector vs Pinecone: Cost, Recall, and FilteringThis comparison examines architecture, filtered recall, operational tradeoffs, and cost per query for self-hosted pgvector and managed Pinecone.
- Qdrant vs Weaviate vs Milvus: Recall, RAM, and ScaleThis comparison covers filtered recall, memory and quantization options, cluster design, multi-tenancy, and scaling across the three engines.
Fundamentals 2 posts
- Cosine Similarity vs Dot Product: How Rankings DifferVector norms determine when cosine similarity and dot product rank results identically and when magnitude changes retrieval order.
- Hybrid Search: BM25, Vector Retrieval, and Score FusionThis guide explains how BM25 and vector retrieval complement each other, why raw scores cannot be combined, and how RRF and alpha fusion work.
Indexing 2 posts
- HNSW M Parameter Tuning: Recall, Memory, and LatencyThe guide explains how M affects graph connectivity and memory, maps engine-specific settings, and shows a sweep for recall@10, p99, and bytes per vector.
- HNSW ef_search Parameter: Recall and Latency TradeoffsThe HNSW ef_search parameter sets query beam width, balancing recall against latency across vector search engines and filtered queries.