Can you guess the platforms that are currently running on a vast majority of the largest global banking, insurance and retail organizations?
Given some industry backlash as of late, it may surprise you that I’m talking about IBM mainframe systems.
Based on a recent survey from BMC, 67 percent of organizations are looking to increase capacity on their mainframe to meet their business priorities. That is to say, they are betting on the mainframe to support their digital future. From my conversations with experts in this space, a variety of details coalesce around two fundamental drivers for this. The first, inferred above, is digital transformation.
As stated at the SHARE conference earlier this year, IDC says organizations are transforming operations to architect a “connected mainframe” that sits at the heart of their new, digital organization, not apart from it.
Sure enough, the mainframe world is awash with web/mobile integration, Java integration, new APIs, all methods of interconnecting z-based systems of record with new front-ends. No longer is the mainframe disconnected, and nowhere is this more prevalent than in the mobile world.
The digital economy has encouraged the creation of a mobile app for virtually every consumer. For example in the banking world, no modern-day user would open a checking account with a bank if they didn’t have a mobile application available too. It’s important to remember that banking applications connect to the mainframe application to receive customer account information in the same way an ATM or web page does. All require access to the mainframe. At a recent industry event, one major global banking client talked about z-based transaction workloads increasing by orders of magnitude, simply because clients now had instantaneous access to their funds (each click or swipe represents a mainframe transaction or two), whereas before this wasn’t happening. Who hasn’t frequently refreshed a banking app while eagerly awaiting a payment?
Another recent example involves an organization that has adopted a mobile front-end to a core mainframe-based insurance system to improve call center efficiency. This simple mashup of mobile and mainframe capabilities has resulted in a twofold increase in throughput of claims, and clients- increasing transactions, but also increasing revenues and customer satisfaction.
The second topic is around the mainframe world witnessing the growth of new workloads. Indeed, new workload is the lifeblood of the mainframe business model. The more CPU cycles consumed, the more the charge. IDC’s prescient commentary at this year’s SHARE conference asserted “workload should run where it should to best support the business.”
Not all workloads are created equal. Linux-based systems, for example, are cheaper and easier to set up. IBM itself has stated that 93% of the top mainframe clients were also running Linux as part of their mainframe setup.
One client of ours was focused on getting to market quickly and branching out into new territories. However, the time to commission the right mainframe system — they required a brand new data center for each new geography — was prohibitively (and necessarily) slow for a z/OS configuration. By contrast, the same system under Linux was a much faster option for them.
Other mainframe organizations have witnessed capacity increases in the z/OS workload, but to avoid reaching the next charging threshold, have moved other non-essential workload to Linux. This was a simple and smart way to manage cost while maintaining a robust level of service. That blend of critical and noncritical, matched against levels of service provided on a mainframe system nowadays, provides a lot of flexibility.
Additionally, there are other server platforms out there that just can’t match the IBM box in terms of value. As discussed at SHARE 2017, U.S. software experts Astadia helped a major banking client move a major IT workload to the IBM z series mainframe. Consolidating other non-IBM mainframes as part of a modernization program, the client simplified their entire datacenter around IBM platforms.
Other issues also mandate extra mainframe cycles. For example, application delivery teams are grappling with making DevOps a mainstream activity for the mainframe, just as it is elsewhere in IT. In such cases, new workload is less significant, and the focus is on refined teamwork, culture and technology. Elsewhere, the growing scrutiny around data breaches means mainframe security teams are busier than ever, but the overall workload impact is limited.
What is clear from customer conversations are the examples of large-scale organizations looking to continue innovation with the mainframe sitting in the middle of their digital transformation initiative.
Given forward-looking statements shared publicly by IBM so far, I’d expect we’ll see IBM unveiling a yet-more-powerful mainframe variant in the coming months. There’s still a lot more workload to manage and, it appears, the mainframe remains very good at it.