MariaDB innovation: vector index performance

Дата публикации: 23-02-2026 16:54:00

Last year I shared many posts documenting MariaDB performance for vector search using ann-benchmarks. Performance was great in MariaDB 11 and this blog post explains that it is even better in MariaDB 12. This work was done by Small Datum LLC and sponsored by the MariaDB Foundation. My previous posts were published in January and February 2025.tl;drVector search recall vs precision in MariaDB 12.3 is better than in MariaDB 11.8Vector search recall vs precision in Maria 11.8 is better than in Postgres 18.2 with pgvector 0.8.1The improvements in MariaDB 12.3 are more significant for larger datasetsMariaDB 12.3 has the best results because it use less CPU per query, This is confirmed by running vmstat in the background.BenchmarkThis post has much more detail about my approach. I ran the benchmark for 1 session. I use ann-benchmarks via my fork of a fork of a fork at this commit.  The ann-benchmarks config files are here for MariaDB and for Postgres.This time I used the dbpedia-openai-X-angular tests for X in 100k, 500k and 1000k.For hardware I used a larger server (Hetzner ax162-s) with 48 cores, 128G of RAM, Ubuntu 22.04 and HW RAID 10 using 2 NVMe devices. For databases I used:MariaDB versions 11.8.5 and 12.3.0 with this config file. Both were compiled from source. Postgres 18.2 with pgvector 0.8.1 with this config file. These were compiled from source. For Postgres tests were run with and without halfvec (float16).I had ps and vmstat running during the benchmark and confirmed there weren\'t storage reads as the table and index were cached by MariaDB and Postgres.The command lines to run the benchmark using my helper scripts are:    bash rall.batch.sh v1 dbpedia-openai-100k-angular c32r128    bash rall.batch.sh v1 dbpedia-openai-500k-angular c32r128    bash rall.batch.sh v1 dbpedia-openai-1000k-angular c32r128Results: dbpedia-openai-100k-angularSummaryMariaDB 12.3 has the best resultsthe difference between MariaDB 12.3 and 11.8 is smaller here than it is below for 500k and 1000kResults: dbpedia-openai-500k-angularSummaryMariaDB 12.3 has the best resultsthe difference between MariaDB 12.3 and 11.8 is larger here than above for 100kResults: dbpedia-openai-1000k-angularSummaryMariaDB 12.3 has the best resultsthe difference between MariaDB 12.3 and 11.8 is larger here than it is above for 100k and 500k
MariaDB innovation: vector index performance appeared first on MariaDB.org

Основное содержимое страницы с новостью.

Схожие новости

#	Наименование новости	Тональность	Информативность	Дата публикации
1	Playing with MariaDB Vector for initial AI tests	0	6.01	28-08-2024
2	Benchmarking vector indexes	0	7.4	27-08-2026
3	The Insert Benchmark vs MariaDB 10.2 to 13.0 on a 24-core server	0	12.6	08-04-2026
4	Big Vector Search Benchmark: 10 databases comparison	0	17.6	11-03-2026
5	Sysbench vs MariaDB on a small server: using the same charset for all versions	0	8.32	06-04-2026
6	MariaDB Vector Support Upstreamed to Open-WebUI: Single-Database RAG Just Got Faster and Simpler	0	11.27	10-06-2026
7	MariaDB Vector in Laravel: insights on choosing an embedding model	0	4.73	18-06-2026
8	Building Smart Semantic Search using PostgreSQL and pgvector. Case Study – Part 1 – Introduction	0	7.25	29-05-2026
9	The Insert Benchmark vs MariaDB 10.2 to 13.0 on a 32-core server	0	6.43	08-04-2026
10	CPU-bound sysbench on a large server: Postgres, MySQL and MariaDB	0	5.53	04-04-2026

Классификация: . Схожих патентов: 0. Схожих новостей: 10. Тональность: 0. Информативность: 8.78. Источник: mariadb.org.