IBM said it is developing a processor chip that can natively execute Arm and IBM Z instructions at the same time. The chip, built on a two‑nanometer process, will contain 11 high‑performance cores running at more than 5.7 GHz and include AI inference accelerators, an on‑chip data processing unit for I/O acceleration, and a large cache for enterprise workloads, with availability targeted for 2028.
Analyst George Cozma of TechInsights said the mainframe market is small in unit terms but remains strategically important because core applications, decades of operational practice, and data gravity make migration expensive and risky. IBM’s blog post noted that the Arm ecosystem includes more than 22 million developers and supports a growing range of cloud‑native, AI, analytics and infrastructure applications. Rocket Software, a vendor of mainframe modernization tools, reported 42 IBM mainframe users in Australia, highlighting the limited but specialized nature of the market.
IBM said the unnamed chip will be used in the upcoming z18 mainframe generation, which follows the z17 model released in 2025 and typically receives a new model every three years. The company said the chip will enable software developers to run Arm‑based and LinuxONE applications alongside traditional Z workloads without requiring separate hardware. Cozma added that the chip represents a move to end the platform’s historical isolation from the broader computing mainstream while preserving mainframe reliability for mission‑critical workloads.
IBM expects the chip to ship in 2028, continuing a three‑year refresh cycle that began with the 2025 launch of the z17. The announcement comes as major cloud providers have introduced their own Arm‑based processors and as customers increasingly adopt Arm‑centric cloud‑native and AI tools.

