Kyndryl: AI interest is driving mainframe modernization projects

Enterprise views on mainframe modernization are shifting as AI becomes more influential. Meanwhile, perennial issues such as security, compliance and skills availability remain top of mind for enterprise mainframe shops, according to Kyndryl, which just released the results of its second annual State of Mainframe Modernization Survey.

Mainframe modernization can take a few forms – enterprise customers can keep data on the big iron, migrate workloads off, or create a hybrid environment. This year, more respondents reported increases in mainframe usage, while fewer are shifting workloads off the mainframe.

“In 2024, more respondents confirmed they are focusing on modernizing on the mainframe or integrating with cloud, and fewer are choosing to move workloads off the mainframe as their primary strategy—a five percentage point drop from last year (30% to 25%). Additionally, 53% of respondents saw their usage on the mainframe increasing this year, with 49% saying their mainframe usage will increase again in the next 12 months,” Kyndryl stated.

In terms of costs, the Kyndryl survey found that mainframe transformationprojects can deliver significant financial benefits for enterprises.

“Respondents undertaking a modernization project reported that the average cost of those projects decreased compared to last year, while profitability of these projects remained the same or even slightly increased,” the study reported.

“Organizations report that modernizing on the mainframe has led to a 9% increase in overall profitability, integrating with other platforms led to a 10% increase in profitability, and moving off the mainframe drove a 12% increase in profits. Additionally, the ROI for these projects is staggering, yielding one year returns between 114% and 225%,” Kyndryl stated.

AI in mainframe environments

Kyndryl also reports greater interest in adopting AI on the mainframe.

AI and generative AI promise to transform the mainframe environment by delivering insights into complex unstructured data, augmenting human action with advances in speed, efficiency and error reduction, while helping to understand and modernize existing applications. Generative AI also has the potential to illuminate the inner workings of monolithic applications, Kyndryl stated.

“Enterprises clearly see the potential with 86% of respondents confirmed they are deploying, or planning to deploy, generative AI tools and applications in their mainframe environments, while 71% say that they are already implementing generative AI-driven insights as part of their mainframe modernization strategy,” Kyndryl stated.

 When asked about the business conditions driving their mainframe modernization investment decisions over the next 12 months, more than a third (36%) named the opportunities offered by AI as a top investment decision, Kyndryl stated.

AI is just beginning to make an impact, however. The majority (80%) of survey respondents said they were in the early or mid-stages of AI integration.

“Businesses are still exploring the possibilities, with over a third (36%) of respondents currently identifying new generative AI use cases,” Kyndryl stated. 

Some aren’t at all convinced of AI’s benefits: 14% of respondents said they don’t plan to use generative AI at all, with security concerns (41%) and other priorities taking precedence (38%). Regulatory concerns (35%) are also a key reason for their reluctance, Kyndryl found.

Security a key factor driving modernization decisions

While AI will likely shape the future for mainframes, a familiar subject remains a key driver for mainframe investments: security.

“Given the ongoing threat from cyberattacks, increasing regulatory pressures, and an uptick in exposure to IT risk, security remains a key focus for respondents this year with almost half (49%) of the survey respondents cited security as the number one driver of their mainframe modernization investments in the year ahead,” Kyndryl stated. 

Two-thirds (66%) of respondents stated that an unparalleled level of security is the most important feature offered by mainframes, and 35% reported that increased security is an advantage of modernizing the mainframe.

Kyndryl found that 92% of respondents factor concerns regarding regulatory compliance (such as with DORA, NIS2, and SEC cyber regulations) into their mainframe modernization strategy.

“Regulatory concerns lead 28% of respondents to modernize more on the mainframe, 45% to integrate with distributed or private cloud environments, and 19% to address regulatory concerns by moving workloads off the mainframe, Kyndryl stated.

Hybrid environments and observability challenges

Another topic for mainframe customers is managing the hybrid environment created by modernized mainframe workloads and cloud environments.

According to the study, 92% of respondents say it is important to have a single dashboard to monitor operations across a hybrid environment, yet 85% find it difficult to do this.

“Survey respondents say such a dashboard would help them oversee infrastructure performance (54%), integration with IT service management (49%) and application performance (45%),” Kyndryl stated.

Increasing visibility can facilitate data collection, event processing, analytics and automation. Such continuous operational improvement enables organizations to gain insights into their software inventory to remove redundancies or automate commands for faster troubleshooting, Kyndryl stated.

Skills deficits remain a challenge

Many respondents still grapple with staffing challenges, and the skills deficit can hinder mainframe modernization projects, Kyndryl found.

• Eighteen percent (18%) of business leaders who are integrating the mainframe with other platforms report that insufficient expertise has been the main challenge to project success.
• More than a quarter (28%) of respondents are concerned that they do not have the right level of skills to effectively modernize their mainframe.
• Fifty percent (50%) of businesses reporting a skills shortage are addressing the issue by hiring or upskilling employees. There is widespread investment in training relating to cybersecurity and regulatory compliance (43% of respondents), and analytics, AI and generative AI (38%).

Kyndryl commissioned Coleman Parkes Research for its study, which surveyed 500 senior leaders at mainframe-based enterprises.

Read more about mainframe modernization

• Kyndryl partnered with Amazon Web Services on services aimed at helping enterprises integrate with or move mainframe applications and data to the AWS cloud.
• IBM recently upped the ante on AI development for the mainframe: The next generation of its Z and LinuxONE mainframes will feature a new processor, the IBM Telum II, which has greater memory and cache capacity than the previous generation, and it integrates a new data processing unit (DPU) specialized for IO acceleration along with enhanced on-chip AI acceleration capabilities.
• Mainframe turns 60 with no plans for retirement: Decades after some predicted its demise, the mainframe is as vital as ever, even in the era of AI.
• IBM program targets mainframe skills shortage: Enterprises continue to hire for new mainframe positions, even as they struggle with recruitment. IBM and its partners are investing in training programs to build up a skilled, sustainable workforce for the mainframe.