The primary Maserati was launched in 1926. The primary Ferrari was launched in 1947. And the primary Porsche was launched in 1948. And my private favourite, the primary Land Rover, additionally was launched in 1948.
What do every of those legendary automobiles have in frequent?
Every predates the mainframe and COBOL, but nobody calls them outdated. Why? As a result of they’ve regularly advanced—embracing trendy engineering, cutting-edge expertise, and innovation whereas sustaining the effectivity, efficiency, reliability, and excellence they had been constructed on. The identical is true for the mainframe.
But, regardless of a long time of steady transformation, some critics nonetheless cling to the parable that mainframes are outdated, inefficient, and unable to combine with trendy IT techniques. This couldn’t be farther from the reality. IBM’s z16, launched in 2023, was constructed for AI, and the z17, resulting from launch this yr, is poised to deal with new workloads with unparalleled safety, scalability, and effectivity. COBOL, the spine of crucial functions, is as straightforward to make use of as any trendy programming language when paired with the correct instruments.
The issue isn’t the mainframe—it’s how we’ve managed and remodeled the functions working on it. As an alternative of strolling away from essentially the most dependable, safe, and high-performing computing platform in historical past, we should always give attention to how it’s evolving to help new workloads, AI-driven insights, and hybrid cloud integration.
A Quickly Modernizing House
The mainframe isn’t standing nonetheless. It’s taking over extra mission-critical workloads than ever, supporting all the pieces from AI-powered fraud detection to high-speed monetary transactions. The truth is, a whopping 72 p.c of the world’s compute runs on mainframes whereas the platform makes up simply 8 p.c of IT prices.
Mainframe transformation includes two issues. First, improvement groups have to harness mainframes’ computing energy, scale, and information storage capabilities. Second, they want these mainframe techniques to align with the automation capabilities that their cousins within the cloud have adopted, making the mainframe software program improvement life cycle extra environment friendly, eliminating handbook processes, and growing the standard and velocity of legacy functions.
DevOps workflows alone gained’t get us there, however instruments are bridging the hole.
With regards to instruments, retailers want mainframe code to be managed identical to cloud or distributed functions, enabling steady integration/steady improvement pipelines, automated testing, and model management whereas sustaining compatibility with legacy environments.
Tradition and the developer expertise additionally play an essential position in mainframe transformation. If the developer expertise for engineers is subpar, a lift to effectivity is unlikely to emerge. Eradicating handbook bottlenecks, decreasing or eliminating context switching, streamlining archaic improvement processes, and adopting an agile tradition are all straightforward methods to enhance the developer expertise.
High quality-Tuning the Mainframe for Authorities Effectivity
Prospects I discuss to typically describe three very completely different—however equally legitimate—paths for fine-tuning their mainframe technique. Some authorities companies select a slow-and-steady method, increasing their mainframe footprint over time as wants evolve. “Our workloads are rising as our inhabitants grows,” one CIO instructed me. “We’re not transferring off the mainframe—we’re rising with it.” For these companies, there’s a pure rhythm of progress that doesn’t require radical change, simply considerate funding as utilization expands.
Others are leaning into modernization by refactoring the code itself. With the assistance of Generative AI-powered code assistants, clients are telling me they’re lastly capable of sort out decades-old functions with confidence. These instruments clarify unfamiliar code in plain language, doc it routinely, and counsel finest practices for making modifications. For presidency groups with restricted entry to senior mainframe builders, this new degree of code intelligence helps bridge the abilities hole and enabling quicker, safer transformation of core functions.
After which there are the companies doubling down—reinvesting within the mainframe by upgrading to the newest zSystems and embracing DevOps practices throughout the board. “If we will do it on the distributed facet, we should always have the ability to do it on the mainframe,” one company chief instructed me. By staying present, these organizations cut back technical debt, help trendy improvement instruments, and guarantee seamless integration into their enterprise-wide DevOps workflows.
Future-Proofing the Mainframe
The builders working with mainframes are additionally enthusiastic about their future. A 2024 Forrester Report discovered that “amongst world infrastructure {hardware} decision-makers, 61% stated that their agency makes use of a mainframe. Of those who use mainframes, 54% indicated that their group would improve its use of a mainframe over the subsequent two years.”
There’s additionally a large ecosystem of distributors constructing instruments to modernize the mainframes.
That’s the reason you see increasingly speak about synthetic intelligence, graphical scanning, and mapping instruments to parse, map, and refactor legacy code bases and monolithic code into extra manageable belongings. AI additionally offers organizations the flexibility to rapidly onboard new sources and get them conversant in their code base quicker to turn out to be extra productive. Builders can pinpoint vital modifications quicker, decreasing planning time and accelerating updates.
These traits are promising, and I don’t doubt that they might enable authorities providers to harness the mainframe’s information storage and processing energy whereas additionally adopting the agility that has been the hallmark of Silicon Valley.