This hardware represents the first major milestone of a strategic collaboration between IBM and Arm, intended to power the next generation of IBM Z and LinuxONE systems.
The development is significant because it integrates the massive ecosystem of Arm-native Linux software directly into the high-security environment used by global banks, insurers, and governments.
By making every core "bilingual," IBM aims to allow enterprises to run modern AI frameworks and cloud-native tools alongside their mission-critical transaction ledgers without the performance penalties of traditional emulation.
According to IBM, this approach provides a "step-function" increase in software compatibility, giving clients access to millions of Arm developers while maintaining the extreme reliability and uptime associated with mainframe infrastructure.
The new silicon is built on a 2-nanometer manufacturing process and operates at frequencies above 5.7 GHz, featuring dedicated accelerators for AI inference and data processing.
While the chip is central to IBM’s strategy to keep its hardware relevant in an AI-driven era, the company indicated that full systems based on this architecture are not expected to ship until approximately 2028.
This long-term roadmap is designed to support "surround workloads" such as real-time fraud detection and document processing, ensuring that newer Arm-based applications can reside on the same memory fabric as sensitive financial data.