MariaDB Introduces Open-Source Vector Preview
MariaDB 11.6 introduces a public preview of its open-source Vector search feature, utilizing the HNSW algorithm to support large language models, aiming to compete with MySQL and attract developers.
Read original article
MariaDB has announced the public preview of its Vector search feature with the release of MariaDB 11.6, aiming to position itself as a leading alternative to MySQL. This initiative is seen as a strategic move to attract developers, especially since Oracle's MySQL primarily reserves new features for its enterprise editions. The upcoming MariaDB 11.6 Vector Edition will include dedicated indexes and functions for vector storage, although it currently lacks a specific vector data type. The implementation utilizes the HNSW algorithm and is designed to support large language models (LLMs). MariaDB's CEO, Kaj Arnö, emphasized the open-source nature of their vector search, contrasting it with Oracle's proprietary offerings. The project has garnered support from various contributors, including AWS. As the demand for generative AI workloads increases, MariaDB is not alone in pursuing vector search capabilities; other platforms like Google and AWS are also enhancing their relational database management systems (RDBMS) with similar features. While the vector search functionality is still in its early stages, it represents a significant development in the database landscape, potentially influencing Oracle's future offerings.
- MariaDB 11.6 introduces a public preview of Vector search, aiming to compete with MySQL.
- The feature is based on the HNSW algorithm and supports large language models.
- MariaDB's vector search is open-source, unlike Oracle's proprietary offerings.
- The project has received contributions from various organizations, including AWS.
- The growing demand for generative AI workloads is driving advancements in vector search across multiple database platforms.
Related
DuckDB: Vector Similarity Search Extension
The vss extension in DuckDB enhances vector similarity search with HNSW indexing for ARRAY columns. Users can optimize queries with distance metrics but should be cautious due to limitations and experimental features.
What's New in MySQL 9.0
Oracle released MySQL 9.0 with new features like Vector datatype and JavaScript Stored Programs. It removes SHA-1 for security and plans regular updates for improved functionalities and security. Users should update regularly.
Is an All-in-One Database the Future?
Specialized databases are emerging to tackle complex data challenges, leading to intricate infrastructures. A universal, all-in-one database remains unfulfilled due to optimization issues and unique challenges of different database types.
SQLite-vec v0.1.0: a vector search SQLite extension that runs everywhere
sqlite-vec v0.1.0 is a new SQLite extension for vector search, supporting multiple programming languages and operating systems. It focuses on brute-force search, with future updates planned for ANN indexing.
SQLite vector search extension that runs anywhere
sqlite-vec is an SQLite extension for fast vector search, supporting float, int8, and binary vectors. It is compatible with multiple platforms and easy to install across various programming languages.
2 commentsI just hope this Vector search won't further fragment the MySQL ecosystem.
Related
DuckDB: Vector Similarity Search Extension
The vss extension in DuckDB enhances vector similarity search with HNSW indexing for ARRAY columns. Users can optimize queries with distance metrics but should be cautious due to limitations and experimental features.
What's New in MySQL 9.0
Oracle released MySQL 9.0 with new features like Vector datatype and JavaScript Stored Programs. It removes SHA-1 for security and plans regular updates for improved functionalities and security. Users should update regularly.
Is an All-in-One Database the Future?
Specialized databases are emerging to tackle complex data challenges, leading to intricate infrastructures. A universal, all-in-one database remains unfulfilled due to optimization issues and unique challenges of different database types.
SQLite-vec v0.1.0: a vector search SQLite extension that runs everywhere
sqlite-vec v0.1.0 is a new SQLite extension for vector search, supporting multiple programming languages and operating systems. It focuses on brute-force search, with future updates planned for ANN indexing.
SQLite vector search extension that runs anywhere
sqlite-vec is an SQLite extension for fast vector search, supporting float, int8, and binary vectors. It is compatible with multiple platforms and easy to install across various programming languages.