AlloyDB for PostgreSQL Columnar Engine Accelerates Vector Search
AlloyDB for PostgreSQL now utilizes its Columnar Engine as a read-optimized, in-memory cache specifically for HNSW vector indexes. This new capability significantly accelerates vector search performance and increases queries per second (QPS) for demanding vector workloads. It primarily benefits developers and architects building applications that rely on high-throughput similarity search. The feature is generally available, providing production-ready support for enhanced vector processing.
Features (1) ›
- AlloyDB for PostgreSQL
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 .
https://docs.cloud.google.com/release-notes#September_21_2026
Related releases
- Google Cloud Spanner Automatically Parameterizes SQL Query Literals for Performance Google Cloud release notes ·
- Gemini Enterprise: Transfer Ownership of Shared Agents Google Cloud release notes ·
- GCP Agent Platform Workbench Enhances Scheduled Notebook Run Options Google Cloud release notes ·
- JupyterLab Client Logs Forwarded to Cloud Logging in Agent Platform Workbench Google Cloud release notes ·
- Agent Platform Workbench M150 Release Updates Upstream Dependencies Google Cloud release notes ·
- Access Context Manager adds extended session length for Workforce Identity Federation in Looker Google Cloud release notes ·