Empowering your AI and cloud data transformation with IBM Db2 12.1.2

5 June 2025

Authors

Satya Krishnaswamy

Director, Hybrid Data Management Development

IBM

Siji Daniel

Senior Product Manager - IBM Db2

IBM

We are excited to announce the General Availability (GA) of IBM Db2 12.1.2, released on 5 June 2025. This release marks a significant milestone in data management designed to power the next generation of AI and cloud-native applications. With cutting-edge features like vector data support, enhanced security, seamless cloud integration and support for open formats, this release empowers enterprises to unlock deeper insights and accelerate innovation. 

What’s New in Db2 12.1.2? 

1. Vector data store: AI-ready data management 

Db2 12.1.2 introduces a new data type for VECTOR data, enabling users to store and manage vector embeddings just like scalar data types. This makes Db2 AI-ready at its core to:

  • Native Vector Storage: Store vector embeddings directly in Db2 tables for efficient data management.
  • Flexible Embedding Models: Generate vectors from any embedding model and store it alongside business data in Db2 tables.
  • Unified Data Management: Seamlessly manage structured and unstructured data together within a single, powerful database platform.

2. Similarity search capability: Powering RAG and AI applications

Db2 now supports semantic similarity search through optimized SQL queries with vector distance metrics. This unlocks transformative AI use cases, such as:

  • Retrieval Augmented Generation (RAG): Enhance large language models with relevant context using nearest-neighbor searches.
  • Recommendation systems: Deliver highly relevant results based on similarity, not just exact matches. For instance, create tailored retail product recommendations based on prior purchases or preferences.
  • Hybrid data queries: Combine traditional SQL with vector operations for rich, context-aware analytics.
  • Fraud detection: Detect in real-time any fraudulent patterns in financial transactions.             
  • Image matching: Find matching images or profiles of people, places or objects in real-time using SQL queries.

These capabilities are accessible using familiar SQL syntax and native Db2 functions - making AI development achievable with speed, simplicity, and security.

3. pureScale Enhancements: Superior performance and flexibility

Db2 12.1.2 delivers significant updates to pureScale to optimize scalability and cost-efficiency:

  • Elastic Fabric Adapter (EFA) on AWS EC2 instances: Leverages low-latency network interfaces for high-performance, transaction-heavy workloads.
  • zLinux sockets support: Enables cost-effective pureScale deployments on IBM Z and LinuxONE servers—ideal for development environments and read-only workloads.
  • zVM support: Enhances virtualization flexibility for optimized hardware utilization in zLinux environments.
  • NVMe I/O fencing (AWS): Accelerates recovery during host failures, ensuring high availability and reliability.

4. Security enhancements: Greater control and compliance

Db2 12.1.2 includes robust security features to meet evolving compliance demands:

  • Option to restrict world-writeable file permissions: Limit world-writable file permissions by allowing users to choose shared group ownership for enhanced control and security.
  • TLS encryption monitoring: New metrics provide insights into the performance impact of secure communications.
  • Mask at read: Applies column masking not just on the result of a query but also when the data is read from the table during query execution, significantly reducing the risk of data inference.

5. Azure BLOB Storage Support: Streamlined cloud integration

Db2 now supports Azure BLOB storage for:

  • External tables
  • Backup, load, and log archival
  • Native Cloud Object Storage (NCOS) tablespaces

Db2 Warehouse on Azure is optimized to benefit from cost-effective, resilient, and scalable object storage solutions—ideal for cloud-native, columnar analytics workloads.

6. Data lake support enhancements: Embracing open formats

Db2 continues its commitment to open data formats by enhancing support for Apache Iceberg: 

  • CRUD (create, read, update, delete) operations fully supported: Added robust UPDATE capabilities for comprehensive data management.
  • Flexible dataset management: Efficiently handle large, distributed datasets in data lake environments.
  • Data accuracy and currency: Keep your data lake consistent and up to date with ease.

Redefine what's possible with your data

Db2 12.1.2 is more than just a version update - it’s a launchpad for next-generation AI applications. Whether you're modernizing legacy systems or pioneering AI-driven, cloud-native solutions, Db2 equips you with the tools you need to innovate with confidence. Explore the possibilities with Db2 12.1.2 today.

Download Db2 Community Edition for free

Latest documentation link