To help organizations unlock the full potential of their data for the AI era, IBM is unveiling IBM Db2 12.1.2, the latest iteration of its flexible, AI-powered database. With enhancements to security, scalability, and support for cutting-edge workloads, Db2 12.1.2 offers an array of new capabilities designed to drive the next generation of AI and cloud applications across every industry.
The first of Db2 12.1.2’s advancements is built-in support for vectorized data, enabling users to store and manage vector data alongside traditional scalar data. This will allow enterprises to more easily power their advanced AI apps, such as retrieval-augmented generation or recommendation systems, backed by the security and trust of a Db2 environment.
The new vector data type also increases the speed and efficiency of similarity searches, which also sees an enhancement in Db2 12.1.2. Now, SQL queries optimized by vector distance metrics better integrate structured and unstructured data, increasing the performance of similarity searches with rich insights. Vector data can be easily utilized with custom built SQL queries and built-in functions, paired with an extension of search beyond exact-matches with contextually linked data points.
Db2 12.1.2 also sees upgrades to pureScale, Db2’s capability for ensuring high availability and scalability. This feature now offers support for low-latency network interfaces with the Elastic Fabric Adapter (EFA) on AWS EC2 instances, serving to optimize performance for more demanding workloads, according to IBM. pureScale on zLinux with sockets for IBM Z and LinuxONE servers help drive cost-efficient configurations, while zVM support offers greater flexibility in virtualizing zLinux environments.
This iteration of Db2 recognizes the need to keep pace with a continuously evolving security landscape, delivering the following enhancements for adhering to compliance standards:
- Restrictable world-writeable permissions on Db2 files and directories enables more granular control over who can write to critical files
- TLS encryption monitoring improvements enable users to track the impact of encryption on performance
- New “Mask at Read” functionality for column masks prevents sensitive data from being inferred during queries
To streamline cloud integration, Db2 12.1.2's remote storage capability now supports Azure BLOB storage, offering support for external tables and utilities such as backup, load utilities, and log archival for users running Db2 on Azure cloud. Additionally, this release continues Db2’s support for open data formats, rounding out full support for CRUD (create, read, update, and delete) operations on Apache Iceberg data lake tables. This functionality helps maintain the agility and freshness of data lakes, making it easy to manage and update data stored in distributed systems, according to IBM.
“Db2 12.1.2.0 brings your database into the future with innovations that support AI, improve cloud performance and enhance security—all while maintaining the reliability and scalability you expect from Db2,” said Miran Badzak, program director, databases, IBM. “As you continue to evolve your data architecture, this release provides the tools you need to stay ahead of the curve.”
To learn more about Db2 12.1.2, please visit https://www.ibm.com/us-en.