IBM has announced the launch of its latest mainframe hardware, the IBM z17, aiming to boost the adoption of artificial intelligence (AI). This device is equipped with the new IBM Telum II processor and supports over 250 AI use cases, including generative AI and AI agents.
Despite the impression that mainframes are outdated, they remain crucial, with 71% of Fortune 500 companies currently utilizing them. According to market research, the mainframe market was valued at approximately $5.3 billion in 2024.
The z17 boasts impressive processing capabilities, handling 450 billion inference operations daily, which is a 50% increase compared to its predecessor, the IBM z16. Designed for seamless integration with various hardware, software, and open-source tools, the z17 demonstrates IBM’s commitment to modernizing its mainframe technology.
Tina Tarquinio, Vice President of Product Management and Design for IBM Z, revealed that the z17 has been under development for five years, predating the recent surge in AI interest that began with the launch of OpenAI’s ChatGPT in November 2022. During this period, IBM engaged in over 2,000 hours of research, gathering feedback from more than 100 customers.
This feedback has proven to align well with the evolving market landscape. Tarquinio emphasized the z17’s adaptability to future developments in AI.
Upon its release, the mainframe will include 48 IBM Spyre AI accelerator chips, with plans to increase this number to 96 within the next year. She highlighted the importance of building headroom for accommodating advancements in AI models, allowing for greater flexibility in resource allocation.
Another notable feature of the z17 is its enhanced energy efficiency. Tarquinio indicated that while the on-chip AI acceleration has increased by 7.5 times, it consumes 5.5 times less energy compared to competitors.
The z17 mainframes will be generally available starting June 8.