IBM has launched the IBM z17, its most advanced mainframe yet and it was built specifically for AI. It combines strong performance, high security, and smart AI tools, all in one platform. Although mainframe seems like some age-old tech, they still have relevance in modern technology.
Over 70% of Fortune 500 companies still use them. They support core systems like banking, healthcare, and logistics. They also power the back end of daily life, and IBM just gave them a major upgrade.
What Makes the IBM z17 Different?
IBM z17 isn’t just a faster machine but a smarter one. The IBM z17 runs on the Telum II processor, designed for speed and AI tasks. It also supports over 250 AI uses, including generative AI and intelligent agents, while using less energy, even as it handles more work.
| Feature | Details |
| Telum II Processor | Built for AI workloads with full encryption |
| 450 Billion Inferences per Day | A 50% increase over the IBM z16 |
| Spyre AI Chips | Starts with 48 chips, scalable to 96 |
| AI Acceleration | 7.5x faster with 5.5x less energy use |
| Seamless Integration | Works with open-source tools and hybrid cloud |
Built for AI Growth
IBM didn’t build this overnight. The company started working on the z17 five years ago, long before the rise of tools like ChatGPT. Tina Tarquinio, VP of Product Management for IBM Z, said the company gathered input from more than 100 clients. In total, IBM spent over 2,000 hours in research and feedback.
She noted something interesting. The feedback they got early on ended up matching where the market is now.
“We didn’t know exactly how AI would evolve,” she said. “But we knew it would need flexibility. That’s what we built.”
Built for AI But Designed to Scale
The z17 doesn’t just run AI, it grows with it. The system includes 48 IBM Spyre AI accelerator chips, and there are plans to double that to 96 chips within a year. That means companies won’t have to swap out their hardware as models get bigger or smarter. The z17 already has room to grow.
“We’re building in AI agility,” Tarquinio explained. “We want this machine to support whatever comes next.”
Cleaner and Greener
AI can use a lot of energy. But IBM made sure the z17 runs efficiently. The machine boosts AI speeds while cutting energy use. Compared to other systems, it uses five and a half times less energy for similar AI tasks. That saves money and supports sustainability goals.
Also read: Is ChatGPT Really Power-Hungry? A New Study Says ‘No’
Coming This June
The z17 will be available globally starting June 8, 2025. It’s built to handle modern workloads while preparing for future demands. It’s also flexible enough to work with tools companies already use, cloud platforms, APIs, and open-source AI.
This isn’t just an upgrade. It’s a statement. IBM believes the future of AI doesn’t just live in the cloud, it also lives in high-performance, secure, on-premise systems like the z17.
