A survey of 1,223 IT professionals who work in mainframe environments finds that 62% work for organizations that have adopted DevOps practices, with 35% relying on DevOps workflows to build and deploy applications on both mainframes and distributed computing systems.
Conducted by BMC, the survey also found that 70% of respondents work for organizations that have embraced artificial intelligence (AI) for IT operations (AIOps), with 33% using it across both mainframes and distributed computing environments. AIOps is now the third most important priority (45%) cited, following compliance/security (64%) and cost optimization (49%).
Additionally, more than three quarters (76%) work for organizations that are to varying degrees making use of generative AI, with 40% already seeing benefits such as increased automation that reduces repetitive tasks (37%), analyzing code and files for vulnerabilities (36%) and business insights (34%).
Finally, the survey finds mainframes are running more diverse workloads, with 64% noting that new mainframe applications are written in Java. A total of 43% added their existing base of Java applications is growing, while 55% are rewriting existing applications in Java.
John McKenny, senior vice president and general manager for Intelligent Z optimization and transformation at BMC, said the survey makes it clear that IT organizations with mainframes are at the forefront of multiple waves of technological innovation. That wave will continue into next year as IBM makes available next-generation platforms that make use of a 5.5 GHz IBM Telum II Processor with eight cores that is better optimized to run artificial intelligence (AI) inference engines along with an IBM Spyre Accelerator chip that makes it simpler to invoke multiple AI models.
Those types of capabilities help ensure that organizations that have invested in mainframes will continue to witness the platform evolve in ways that can, for example, efficiently run Python applications, added McKenny. In fact, 94% of survey respondents continue to have a positive view of the mainframe, he noted.
Make Mainframes More Accessible
The goal now is to make mainframes more accessible to a wider range of IT professionals at a time when many of the individuals who currently have that expertise are getting ready to retire, noted McKenny. Rather than requiring the next generation of IT teams to learn how to master the nuances of the mainframe, the tools required to manage mainframes are being modernized for the AI era, he added.
Of course, the cost of acquisition of a mainframe is still higher than other platforms but given the way how IBM bills organizations for usage of mainframes the platform over time continues to provide in the final analysis a lower total cost of ownership (TCO) by, for example, enabling some workloads running on LinuxOne to run at no additional cost. The challenge now is making sure there is still enough mainframe expertise available for organizations to achieve that goal.
Each organization will ultimately need to decide which type of workload makes the most sense to run on what type of platform for themselves. However, the mainframe continues to run certain classes of online transaction processing (OLTP) applications and associated analytics at levels of performance that other platforms still can’t match.