IBM has unveiled its latest generation of mainframe, the z17.
Engineered specifically to support AI capabilities across hardware, software, and systems operations, the company said the z17 is the result of five years of design and development.
Powered by a second-generation on-chip AI accelerator built on the IBM Telum II processor, the company claims the z17 has the ability to process 50 percent more AI inference operations per day than its predecessor, the z16.
As a result, the company says the new mainframe features increased frequency and compute capacity alongside a 40 percent growth in cache, enabling more than 450 billion inferencing operations in a day and one-millisecond response time.
The z17 also features the 32-core IBM Spyre accelerator to provide additional AI compute capabilities alongside the Telum II.
Due to be available in Q4 2025 via a 75-watt PCIe adapter, IBM says the Spyre Accelerator has been “specially engineered to bring generative AI capabilities to the mainframe,” including running AI assistants such as IBM watsonx Code Assistant for Z and IBM watsonx Assistant for Z, and leveraging enterprise data contained in the system.
First announced at 2024's Hot Chips event, each Telum II features eight cores running at 5.5GHz, with 36MB L2 cache per core for a total of 360MB. The virtual level-4 cache of 2.88GB per processor drawer provides a 40 percent increase over the previous generation.
Meanwhile, each Spyre features up to 1TB of memory, built to work in tandem across the eight cards of a regular IO drawer, and designed to consume no more than 75W per card. Each chip will have 32 compute cores supporting int8 and fp16 data types.
IBM announced the original Telum chip in 2021, and the z16, launched in 2022, was the first to feature the processor.
The new mainframe will run on z/OS 3.2, the updated version of the operating system for IBM Z, planned to be released in Q3 2025, and IBM says the inclusion of the tenth generation of IBM Storage DS8000 will allow users access to critical workloads in addition to “consistent and optimized” data performance.
“The industry is quickly learning that AI will only be as valuable as the infrastructure it runs on,” said Ross Mauri, general manager of IBM Z and LinuxONE, IBM. “With z17, we’re bringing AI to the core of the enterprise with the software, processing power, and storage to make AI operational quickly. Additionally, organizations can put their vast, untapped stores of enterprise data to work with AI in a secured, cost-effective way.”
The IBM z17 will be generally available on June 18, 2025.
