IBM Z and LinuxONE Dual-Architecture Processor

IBM has unveiled a dual-architecture processor designed for future IBM Z and LinuxONE systems, enabling the native execution of both Arm and IBM Z/LinuxONE instructions on a single processor core. This development allows enterprises to run Arm-native Linux environments alongside z/OS and Linux on IBM Z, integrating the broad Arm software ecosystem into IBM's high-security, mission-critical mainframe infrastructure.

Unified Core Architecture

The new processor does not utilize separate cores for different architectures. Instead, each individual core is architected to natively execute both Arm and IBM Z (or LinuxONE) instructions concurrently.

According to community analysis, the cores decode and execute both s390x and Arm AArch64 instructions, dynamically switching between the two ISA modes and converting them into micro-ops, with mode switching managed by the hypervisor.

Hardware Specifications

The processor is built on a 2 nanometer technology node and features the following technical specifications:

  • Clock Speed: High-performance cores operating at more than 5.7 GHz.
  • Core Count: 11 high-performance cores per chip.
  • Cache: A large cache architecture optimized for demanding enterprise workloads.
  • AI Acceleration: Integrated AI inference accelerators specifically designed for in-transaction fraud detection.
  • I/O Acceleration: A dedicated on-chip data processing unit (DPU) for I/O acceleration.

Enterprise Integration and Ecosystem Expansion

This processor is the first major milestone of a strategic collaboration between IBM and Arm established in April 2026. By bringing Arm natively to the mainframe, IBM aims to provide organizations with greater choice in how they deploy AI and cloud-native applications while maintaining the established security, reliability, and scale of the IBM Z platform.

Arm workloads on this platform will benefit from IBM's enterprise-grade capabilities, including:

  • Hardware-level fault detection and recovery.
  • Advanced encryption and secure key management.
  • Enhanced reliability and availability characteristics.

Community Perspectives

Technical observers have noted the historical context of this move, comparing it to previous IBM initiatives like eCLipz, which sought to unify z and POWER architectures.

"20 years ago, IBM had their eCLipz initiative to unify z and POWER. What actually shipped was two different CPUs sharing some design... looks like they finally shipped it, just replacing POWER with ARM"

Other discussions focused on the implementation details of the ISA switching, with some questioning if the design is conceptually similar to hardware-based code translation or previous architectural experiments like the Data General Fountainhead project.

Sources

Related

  • Dispatch
    Apple M6 and M5 Ultra launch: 2 nm silicon, quad‑die architecture, and massive AI compute

    Apple unveiled the M6 2 nm chip in a new Mac mini and the M5 Ultra quad‑die SoC in a new Mac Studio, delivering up to 2.4× faster CPU performance, 30% more AI GPU compute, and up to 1.2 TB/s memory bandwidth for on‑device large language models.

  • Dispatch
    IBM 0.7nm Chip Technology: Nanostack 3D Architecture and the Angstrom Era

    IBM has unveiled a sub-1 nanometer chip technology at the 0.7nm (7 angstrom) node, utilizing a new 3D 'nanostack' architecture to nearly double transistor density over its 2nm node.

  • Project
    zml/zml

    A production inference stack that decouples AI workloads from proprietary hardware, allowing a single codebase to run models on NVIDIA, AMD, Intel, and TPU/Trainium accelerators.

  • Dispatch
    OpenAI and Broadcom Unveil Jalapeño LLM Inference Chip

    OpenAI and Broadcom have introduced Jalapeño, a custom-designed AI accelerator optimized specifically for LLM inference to improve performance per watt and reduce compute costs.

  • Project
    Samsung/ONE

    ONE is a high-performance on-device neural network inference framework that enables AI models from TensorFlow and PyTorch to run efficiently across various hardware accelerators.