You can now use the AlloyDB Columnar Engine as a read-optimized, in-memory cache for HNSW vector indexes
GAVerifiedAdded Sep 22, 2026
You can now use the AlloyDB Columnar Engine as a read-optimized, in-memory cache for HNSW vector indexes. This feature is generally available ( GA ). It accelerates vector search performance and increases queries per second (QPS) for vector workloads. For more information, see Accelerate queries with the Columnar Engine .
Topics: Vector search, Performance
Summaries of vendors' own notes. Product names and logos belong to their owners; logos via logo.dev.
More AlloyDB releases
Every AlloyDB release| Date | Release | Type |
|---|---|---|
| Sep 11 | AlloyDB audit log pipeline status available in Cloud Monitoring Update | Update |
| Sep 2 | AlloyDB real-time BigQuery access adds limit pushdown and runtime projects Preview | Preview |
| Aug 17 | MCP Toolbox for Databases adds AlloyDB observability and query insights Update | Update |
| Aug 7 | AlloyDB can sync BigQuery tables in preview Preview | Preview |
| Aug 7 | AlloyDB integration with BigQuery enters preview Preview | Preview |
| Aug 4 | AlloyDB now supports Best Matching 25 (BM25) indexes for full-text search in Preview Preview | Preview |
| Jul 28 | Note: IAM group authentication is available in Preview for new AlloyDB clusters Preview | Preview |
| Jul 27 | AlloyDB write endpoints are now available in Preview Preview | Preview |
