Skip to content
Tech News
← Back to articles

IBM is building a dual-architecture mainframe chip that will run Arm and Z software on the same cores

read original more articles
Why This Matters

IBM's development of a dual-architecture mainframe chip that natively supports both Arm and Z software marks a significant advancement in mainframe technology, enabling seamless integration of cloud, AI, and transaction workloads on a single system. This innovation could reshape how enterprises deploy diverse applications, combining the strengths of Arm's ecosystem with IBM's mainframe capabilities. The move highlights IBM's commitment to evolving mainframe architecture to meet modern computing demands and expand its relevance in the cloud and AI era.

Key Takeaways

First look: IBM is developing a mainframe chip designed to run Arm software directly alongside IBM Z and LinuxONE code. The approach could let customers add Linux, cloud and AI applications while keeping core transaction workloads on the same system.

The processor was showcased at Hot Chips 2026 and is expected to use 11 IBM Z cores built on a 2-nanometer process. IBM said the cores will operate above 5.7 GHz and support native execution of both z/Architecture and AArch64 instructions.

That's the central change here: IBM isn't placing separate Arm and Z cores on the same chip, and it isn't using emulation. Instead, Arm instruction support is being built directly into the mainframe core, with each processor core able to run Arm instructions concurrently with IBM Z or LinuxONE instructions.

The design could make it easier for customers to run Arm-based Linux, cloud-native and AI software alongside mainframe applications that handle core transaction and database workloads. IBM said it wants to combine the reach of the Arm software ecosystem with the scalability, security and high-availability features associated with Z and LinuxONE.

The chip will also include a set of specialized accelerators. IBM plans AI inference hardware for in-transaction fraud detection, a data processing unit for I/O acceleration, and separate accelerators for compression, cryptography and sorting.

Its cache system is substantial: IBM disclosed a 36 MB private L2 cache, a 432 MB virtual L3 cache and as much as 3.5 GB of virtual L4 cache, resources intended to keep latency low for large databases and transaction-heavy workloads.

IBM hasn't named the processor or announced the system that will use it, but the company said it expects the design to arrive in 2028. IBM launched its current z17 mainframe generation in 2025 and has generally updated the platform on a three-year schedule, making the chip a likely candidate for a future z18 family.

The announcement comes as Arm gains a larger role in enterprise infrastructure. Major cloud providers have introduced their own Arm processors, while open-source infrastructure, AI and cloud-native software increasingly supports the architecture. According to Meredith Stowell, IBM's vice president for the zSystems Ecosystem, the Arm ecosystem includes more than 22 million developers and a growing selection of cloud-native, AI, analytics, infrastructure and enterprise software.

IBM's mainframes remain central to many large organizations, particularly in financial services. IBM says its Z and LinuxONE systems run 68% of the world's production IT workloads, while 77 of the world's 100 largest banks use the platforms.

Still, the mainframe market is limited compared with the broader server market. That can make IBM Z a lower priority for independent software vendors, even as the systems continue to run some of the most critical workloads inside major companies.

... continue reading