Last time we covered Db2 12 for z/OS Function Levels and how that was the basis for Db2v13.
Another synergy with the z Systems hardware is that Db2 13 for z/OS takes advantage of the new IBM Z Telum processor for AI workloads. And AI capabilities are one of the big new capabilities of Db2 13 for z/OS.
Perhaps the most important AI-related new feature set in Db2 13 for z/OS is SQL Data Insights. This refers to the delivery of new AI functions built into Db2. By combining deep learning in AI with the new IBM Z processor, SQL Data Insights enables users to write SQL-based semantic queries on their Db2 tables and views.
Because it is part of Db2, there is no need to move data around (such as with ETL) before you can perform AI functions on it. The data lives on the mainframe, Db2 lives on the mainframe, and the AI delivered with SQL Data Insights is part of Db2!
From a developer perspective, the basic functionality of SQL Data Insights is delivered via three new built-in functions:
- AI_SIMILARITY—computes a score to enable comparing data for its similarity. For example, you could specify a customer and ask Db2 to return other customers that are most similar to it
- AI_SEMANTIC_CLUSTER—computes a semantic clustering score of a member argument against a set of clustering arguments. For example, you could specify a set of customers and ask Db2 to return other customers that best belong to that set
- AI_ANALOGY—computes an analogy score between two sets of values. This function works like an analogy: for example, A is to B as C is to ?
SQL Data Insights should be helpful to organizations looking to uncover heretofore unknown relationships in their data. If you think about these new functions, it would be very difficult to code the same capabilities into your applications without them. But because SDQL Data Insights uses built-in functions, you can use it anywhere you use SQL!
Of course, this is a high-level overview of SQL Data Insights. Consult the documentation for the actual formulation of queries using these functions. And it is important to understand, before these functions can be used, you need to train the model for AI by collecting key statistics and building metric scores for the functions to use. And as anybody who has built models for AI knows, this process can be lengthy and consume a lot of CPU resources. Fortunately, zIIPs can be used to build the models, thereby minimizing the cost of training.
With SQL Data Insights, Db2 13 enables you to extend your queries into the realm of AI, which is exciting because it can help you to gather heretofore undiscovered insight into your data. But there are other AI capabilities delivered with Db2 13 in addition to SQL Data Insights.
AI is infused into Db2 13 delivering improved performance, reduced cost, and more proactive system monitoring. The Db2 engine code can intelligently adapt its behavior as it learns from processing your workload. Consider for example, the index lookaside improvements introduced in Db2 13. By storing additional details of index access and Db2 can enable more lookaside operations for more indexes and use cases. Db2 13 index lookaside will minimize root-to-leaf index for INSERT, UPDATE, and DELETE operations regardless of the index cluster ratio. This can reduce index GETPAGE requests, particularly for modification-heavy workloads.
Additionally, Db2 13 helps to simplify building models for AI and exploits the speed of the IBM z16 for training and querying data. Of course, AI is not the only new functionality delivered in Db2 13 for z/OS. These are the business-as-usual features that IBM continues to introduce to make Db2 more usable and efficient. These Db2 13 features can be broken down into the following six categories:
- Application Support
- Availability, Resiliency, and Scalability
- Simplification and Serviceability
Next time, we’ll look at these in more detail.
Regular Planet Mainframe Blog Contributor
Craig Mullins is President & Principal Consultant of Mullins Consulting, Inc., and the publisher/editor of The Database Site. Craig also writes for many popular IT and database journals and web sites, and is a frequent speaker on database issues at IT conferences. He has been named by IBM as a Gold Consultant and an Information Champion. He was recently named one of the Top 200 Thought Leaders in Big Data & Analytics by AnalyticsWeek magazine.