General Manager and Editorial Director,
Techstrong Group

One of the issues that digital business transformation brings to the fore is what to do with legacy applications that reside on mainframe platforms that drive trillions of transactions. For several years now, a debate has raged over to what degree organizations should either replace these applications, migrate them to a cloud service or modernize them by adding application programming interfaces (APIs) to modernize them. The argument for replacing or migrating these applications by, for example, encapsulating them in containers, usually comes down to whether an organization can find and retain the talent required to maintain a mainframe platform.

Modernizing mainframe applications is, conversely, a less complicated endeavor that enables organizations to leverage investments in transaction processing and associated analytics applications they have already made. The challenge is that building, deploying, maintaining and securing all the APIs required also requires a significant amount of effort.

However, when all things are considered, that latter approach enables organizations that have invested in mainframes to drive digital business transformation initiatives much faster, argues Barry Baker vice president of product management for IBM Z mainframes at IBM. In fact, given all the time, effort and expense required to migrate a mainframe application, such approaches to modernization present organizations with an overly simplistic view of what modernization should be all about, says Baker.

Instead, the primary goal should be to make it simpler for other applications to leverage a mainframe platform that six decades after its initial launch not only still outperforms the transaction processing capabilities of any other platform, but is also being advanced with the addition of processors that are optimized for artificial intelligence (AI) workloads. IBM, in the first half of 2022, is on track to deliver a Telum processor for Z-Series mainframes that is optimized for processing the inference engines that AI models employ to optimize transaction processing applications. That capability, for example, will make it possible to identify instances of fraud in real time.

IBM, as part of an effort to make it simpler to modernize mainframe applications, has also launched an IBM Z and Cloud Modernization Center through which it is making available assets and methodologies co-creation with clients via IBM Consulting and third-party partners such as Capgemini and Deloitte Consulting. The goal is to create a unified portal that makes it simpler for IT organizations to discover best practices for modernizing mainframe applications, says Baker.

There is certainly no doubt that going into 2022 the debate over whether to maintain and expand investments in mainframes will only intensify. Every major cloud service provider already has some form of a mainframe migration initiative in place. There are, of course, always going to be workloads that might make more sense to run on a public cloud. However, it’s also arguable that most of the mainframe applications that lend themselves to run on another platform have already been migrated or retired. The workloads that run on the mainframe today do so because the senior leadership of most organizations has already determined that approach best suits their requirements. The challenge and the opportunity now will be determining how best to integrate those workloads within the context of a larger digital business transformation strategy that, by definition, spans the entire enterprise.