Qdrant Internals - Qdrant (original ) (raw ) Qdrant InternalsTake a look under the hood of Qdrant’s high-performance vector search engine. Explore the architecture, components, and design principles the Qdrant Vector Search Engine is built on.
Built for Vector SearchWhy add-on vector search looks good — until you actually use it.Evgeniya Sukhodolskaya & Andrey VasnetsovFebruary 17, 2025
Introducing Gridstore: Qdrant's Custom Key-Value StoreWhy and how we built our own key-value store. A short technical report on our procedure and results.Luis Cossio, Arnaud Gourlay & David MyrielFebruary 05, 2025
Qdrant Internals: Immutable Data StructuresLearn how immutable data structures improve vector search performance in Qdrant.Andrey VasnetsovAugust 20, 2024
Vector Search as a dedicated serviceWhy vector search requires a dedicated service.Andrey VasnetsovNovember 30, 2023
Google Summer of Code 2023 - Polygon Geo Filter for Qdrant Vector DatabaseA Summary of my work and experience at Qdrant's Gsoc '23.
Binary Quantization - Vector Search, 40x FasterBinary Quantization is a newly introduced mechanism of reducing the memory footprint and increasing performanceNirant KasliwalSeptember 18, 2023
Qdrant under the hood: io_uringSlow disk decelerating your Qdrant deployment? Get on top of IO overhead with this one trick!
Product Quantization in Vector Search | QdrantDiscover product quantization in vector search technology. Learn how it optimizes storage and accelerates search processes for high-dimensional data.Kacper ŁukawskiMay 30, 2023
Scalar Quantization: Background, Practices & More | QdrantDiscover the efficiency of scalar quantization for optimized data storage and enhanced performance. Learn about its data compression benefits and efficiency improvements.Kacper ŁukawskiMarch 27, 2023
Minimal RAM you need to serve a million vectorsHow to properly measure RAM usage and optimize Qdrant for memory consumption.Andrei VasnetsovDecember 07, 2022
Filtrable HNSWHow to make ANN search with custom filtering? Search in selected subsets without loosing the results.Andrei VasnetsovNovember 24, 2019