2025 was a year of momentum for the mainframe ecosystem.

Across hardware, software, workforce development, and community engagement, the platform continues to evolve in ways that directly reflect how enterprises are building secure, hybrid, and AI-enabled environments today. From major infrastructure investments and operating system releases to expanded education efforts and longstanding community milestones, last year showcased where the mainframe is headed and its important role in enterprises across industries.

To capture that momentum, we’re counting down 10 of the most consequential mainframe moments of 2025 as identified by SHARE’s Editorial Advisory Committee.

10. IBM Backs the Platform With Long-Term U.S. Investment

In April, IBM announced an investment of $150 billion in the United States over the next five years, including more than $30 billion in research and development to advance and continue U.S.-based manufacturing of mainframe and quantum computers. According to the announcement, today, more than 70% of the world’s transactions by value run on IBM mainframes manufactured in the United States, and this investment reinforces IBM’s commitment to sustaining and expanding that role.

9. Market Research Continues To Show Mainframe Growth

A Mordor report noted that in 2025 the mainframe market was valued at USD 5.33 billion and estimated to reach USD 7.54 billion by 2031. Market research models continued growth in the mainframe market, with forecasts tied to enterprise demand for transaction processing, security, and hybrid deployment models.

Across reports, analysts highlighted consistent drivers, particularly in financial services and government, where transaction throughput, data integrity, and operational continuity remain central requirements. Industry research from International Data Corporation, research and advisory firm ISG, and others continues to point to these sectors as the primary anchors for large‑scale mainframe workloads.

8. Workforce Development Moves to the Forefront

Workforce development emerged as a defining issue for the mainframe ecosystem. Industry reporting increasingly focused on the challenge of sustaining deep platform expertise as experienced practitioners retire and enterprise reliance on the mainframe continues. Kyndryl’s 2025 survey data underscored the reality that organizations are struggling to secure and retain the talent they need, with “70% of organizations … struggling to find the right multiskilled talent needed” to support their environments.

Upskilling is noted as a key for companies navigating retirements. The same is true for professionals looking to stay relevant. “You risk falling into the mainframe skills chasm if you’re not actively learning,” notes a Planet Mainframe article. 

That perspective is echoed within the SHARE community. On SHARE’d Intelligence, professionals reflected on ways mainframes can expand their knowledge through upskilling — “working in the mainframe environment requires knowledge not only of the platform, but also of its connected components and applications.”

Those workforce pressures were increasingly reflected upstream in education and talent pipeline efforts. In 2025, universities, industry partners, and user groups continued to collaborate on structured mainframe education pathways.

Announcing the IBM Mainframe Skills Council, IBM emphasized the importance of preparing the next generation, stating that “modern mainframe skills are critical as businesses implement AI, modernize their applications, integrate across the enterprise and continue digital transformation.” The council brings together universities, employers, and the community to align academic programs with real‑world enterprise roles, reinforcing higher education’s role in sustaining mainframe expertise.

7. Open Mainframe Project Marks 10 Years of Community Impact

The Open Mainframe Project celebrated its 10th anniversary, closing a decade of sustained open‑source collaboration across the mainframe ecosystem. Founded in 2015 under the Linux Foundation, the project was established to provide a neutral home for shared tooling, open development practices, and cross‑vendor cooperation — an approach that was still emerging in the mainframe world at the time.

Over the past 10 years, the Open Mainframe Project has helped normalize open frameworks, community testing, and shared education as part of everyday mainframe work. What began as an effort to improve access to Linux and open source on the platform has expanded into a broad portfolio of projects spanning developer experience, automation, education, and language sustainability, including Zowe, Galasa, Mainframe Open Education, and the COBOL programming course.

6. AI Tools Start Paying Real Dividends for Developers

AI-assisted tooling on IBM Z began delivering tangible value in everyday development workflows, particularly around testing, code comprehension, and operational efficiency. One clear inflection point came with the expansion of IBM Test Accelerator for Z, which IBM described as a solution that “simplifies z/OS application testing and accelerates quality goals” by enabling automated unit testing, early integration testing, and comprehensive integration testing.

IBM also emphasized its ability to “automatically generate test data for challenging program paths, significantly boosting code coverage and accelerating testing processes,” directly addressing longstanding constraints around manual test creation and validation in mainframe environments. Assisted tooling on IBM Z began delivering tangible value in everyday development workflows, particularly around testing, code comprehension, and operational efficiency.

Beyond testing, AI tools increasingly supported how developers and operators understand and manage complex systems. IBM positioned watsonx Code Assistant for Z as a way to help teams “quickly pinpoint where business logic resides to simplify and streamline application enhancements,” reinforcing AI’s role in comprehension and safe change rather than wholesale code generation. At the same time, AI‑driven operations tooling, such as IBM Z Operations Unite, focused on reducing cognitive overload, with IBM describing its purpose as “unifying workflows, prioritizing events, and assessing business impact” to help teams make faster, more confident decisions.

5. Conferences and Community Learning Drive Ecosystem Expansion (SHARE, TechXchange, and More)

2025’s major gatherings — SHARE events, IBM TechXchange, and other ecosystem conferences — kept the focus on what teams are working on right now: AI, security, hybrid architecture, operations, and workforce development.

SHARE’s on-the-ground coverage from Cleveland captured that mix of deep technical content and strategic conversation, while broader conference reporting highlighted how quickly AI, governance, and operational readiness have moved to the center of enterprise planning.

Community learning also extended beyond conference halls. Alongside formal events, year‑round educational efforts increasingly focused on reaching audiences outside the traditional mainframe community.

One example was Broadcom’s Big Iron Bits video series, which was highlighted in SHARE’d Intelligence as an intentionally approachable way to explain mainframe concepts to business leaders, new technologists, and other non‑mainframe stakeholders. The series, which has gone viral, was designed less for technical training and more as a conversation starter — helping mainframers articulate platform value in language that resonates beyond the core ecosystem.

4. SHARE Marks 70 Years of Collaboration and Technical Depth

SHARE celebrated its 70th anniversary, a milestone that highlights the organization’s long-running role in collaboration, education, and community-driven progress. The anniversary celebration on SHARE’d Intelligence focused less on nostalgia and more on continuity: how the same “learn, connect, share” engine drives today’s work across security, systems, operations, and emerging technologies.

3. zPDT Sunsetting Triggers Ecosystem-Level Conversations About Dev/Test Access

IBM’s decision to sunset zPDT created a concentrated conversation in the independent software vendor (ISV) ecosystem about development and testing access models, cost predictability, and capability parity for replacement options.

SHARE engaged directly — bringing together ISV perspectives and documenting dialogue with IBM, including IBM’s stated motivations and the ongoing work toward replacement pathways.

2. z/OS 3.2 Arrives, Ready for AI and Hybrid Cloud

The z/OS 3.2 release delivered updates aligned with current enterprise priorities, particularly around AI enablement, hybrid cloud integration, and operational efficiency.

Key highlights included:

• Hardware-accelerated AI support, designed to take advantage of new processor abilities
• Hybrid cloud integration, enabling more flexible access to z/OS data and services
• Improved security, including continued progress toward quantum-safe capabilities
• Simplified system management and modernization tools to reduce operational friction

These enhancements helped organizations more fully leverage the capabilities of the latest IBM Z hardware and integrate z/OS environments into broader enterprise architectures.

1. IBM z17 Launches as the First AI-Native Mainframe

At the top of the list is the IBM z17.

Announced and rolled out in 2025, z17 was engineered with AI capabilities integrated across hardware and software, reflecting how enterprises are embedding AI directly into mission‑critical workflows.

Key elements of the z17 platform include:

• The Telum II processor, featuring on‑chip AI acceleration for real‑time inference
• The Spyre AI Accelerator, designed to support scalable generative and agentic AI workloads
• Tight alignment with z/OS 3.2 and VSEⁿ 6.4, enabling coordinated platform adoption

Rather than standing alone as a hardware refresh, z17 arrived as part of a broader step forward for the platform — supporting AI, security, and large‑scale transaction processing in place, where enterprise data already resides.