IBM is launching several tools and features for its IBM Z mainframes. This is intended to give customers who rely on mainframe architecture access to modern AI capabilities.

Modern (generative) AI seems to exist at the mercy of the cloud, but IBM wants nothing to do with it. By introducing new functionalities, IBM wants to ensure that customers of its mainframe computers can also get started with modern AI applications and, in particular, the insights resulting from them.

Five new abilities

IBM is therefore providing five innovations that increase the AI ​​capabilities of the IBM Z mainframes:

• AI Toolkit for IBM Z and LinuxOne: This toolkit supports popular frameworks like TensorFlox, so enterprises can start using it on their mainframes. Availability is planned for the end of the year.
• Python AI Toolkit for IBM z/OS: This toolkit contains libraries that support AI and ML, but also meet IBM’s own security and privacy requirements.
• Machine learning for IBM z/OS (Enterprise and Core editions): This capability comes in the form of a full-fledged platform that enables organizations to build, deploy, manage, and operate ML models on z/OS. IBM is based on watsonx.ai. The focus of the toolkit is transparency, so that the results of trained models can also be explained.
• Cloud Pak for data on IBM Z: This AI tool can automate the construction of ML models. Users need to upload data, specify their problem and some operational parameters, and the solution will automatically work towards the desired result.
• AI-powered IBM z/OS 3.1: The latest version of IBM z/OS will be generally available on September 29th. The system has AI system services on board that allow it to optimize IT processes, simplify administration, improve performance and increase accessibility. IBM believes that users need less specialized knowledge to get started.

IBM wants to take care of the entire lifecycle of AI and is in an interesting position to do so. The company is certainly the standard for mainframes, but IBM was also good at AI. The generative AI hype now makes AI almost synonymous with OpenAI and ChatGPT, but don’t forget that not long ago artificial intelligence was mostly reminiscent of IBM Watson. IBM’s current AI capabilities are built on this.