IBM and Arm Partner to Run Arm Software on Mainframes for AI Workloads

IBM and Arm are collaborating to enable Arm-based software to run on IBM's Z and LinuxONE mainframes, combining enterprise-grade reliability with Arm's power-efficient computing for AI and data-intensive workloads.

The Partnership

The collaboration focuses on three key areas:

1. Virtualization for Arm on Mainframe

Using virtualization to allow Arm-based software environments to operate within IBM's enterprise computing platforms.

2. AI and Data-Intensive Workloads

Getting enterprise systems to recognize and execute Arm applications, with the goal of fitting Arm-based environments into enterprise-grade reliability and security requirements.

3. Long-term Ecosystem Growth

Creating shared technology layers between platforms for greater flexibility in application deployment and management.

Strategic Rationale

IBM sees this as enabling enterprise customers to:

• Run the latest AI tools and applications alongside mission-critical workloads
• Maintain existing mainframe investments while adopting new architectures
• Benefit from Arm's broad software ecosystem and power efficiency

Why It Matters

This is a significant shift for IBM's mainframe business. Traditionally locked into its own instruction set architecture (IBM Z), the company is now embracing Arm compatibility — a recognition that the AI ecosystem is overwhelmingly Arm-based.

Mainframes still process a massive share of the world's financial transactions, and bringing Arm software compatibility could unlock new use cases in banking, insurance, and government sectors where mainframes dominate.

Context

• IBM Z mainframes process an estimated 90%+ of credit card transactions globally
• The AI model ecosystem has largely standardized on Arm and NVIDIA architectures
• This move positions IBM's mainframes as AI-capable rather than legacy systems