IBM introduces Spyre AI accelerator to market
Cette publication existe aussi en Français
IBM has unveiled its new Spyre Accelerator, a custom AI chip designed to boost generative and agentic AI workloads in enterprise systems. The accelerator will become commercially available on October 28 for IBM z17 and LinuxONE 5 systems, with availability for Power11 servers following in December.
For eeNews Europe readers — particularly those following developments in enterprise AI hardware, system integration, and semiconductor design — this launch is significant. Spyre represents a tangible bridge between mainframe reliability and cutting-edge AI acceleration, offering insight into how AI processing is moving closer to core business workloads.
Bridging AI and mission-critical workloads
The Spyre Accelerator was developed to enable low-latency inferencing and real-time responsiveness for AI agents running alongside traditional enterprise applications. Built with 32 accelerator cores and 25.6 billion transistors using a 5nm process, each Spyre chip delivers high throughput while maintaining enterprise-grade security and energy efficiency.
IBM says customers will be able to cluster up to 48 Spyre cards in an IBM Z or LinuxONE system, or 16 cards in a Power11 server, providing scalable AI performance on-premises. This on-prem approach, IBM notes, allows clients to keep sensitive data secure while running large-scale generative and agentic AI workloads.
“One of our key priorities has been advancing infrastructure to meet the demands of new and emerging AI workloads,” said Barry Baker, COO, IBM Infrastructure & GM, IBM Systems. “With the Spyre Accelerator, we’re extending the capabilities of our systems to support multi-model AI – including generative and agentic AI. This innovation positions clients to scale their AI-enabled mission-critical workloads with uncompromising security, resilience, and efficiency, while unlocking the value of their enterprise data.”
From research to commercialization
Spyre’s origins trace back to the IBM Research AI Hardware Center, where it began as a prototype designed to explore energy-efficient AI computation. Over several development cycles and collaborations — including with the University at Albany’s Center for Emerging Artificial Intelligence Systems — the chip evolved into a production-ready system-on-chip suitable for enterprise environments.
“We launched the IBM Research AI Hardware Center in 2019 with a mission to meet the rising computational demands of AI, even before the surge in LLMs and AI models we’ve recently seen,” said Mukesh Khare, GM of IBM Semiconductors and VP of Hybrid Cloud, IBM. “Now, amid increasing demand for advanced AI capabilities, we’re proud to see the first chip from the Center enter commercialization.”
Enterprise AI, closer to the data
For mainframe users, Spyre complements IBM’s Telum II processor, enabling near real-time inferencing for applications such as fraud detection, retail automation, and predictive analytics. On IBM Power systems, Spyre integrates with an AI services catalog that allows one-click installation of enterprise AI workflows, supporting seamless data conversion and integration for large-scale generative AI.
With the launch of Spyre, IBM is positioning itself at the intersection of AI acceleration and enterprise infrastructure, signaling a shift where generative AI no longer sits in the cloud alone — but increasingly, at the heart of on-prem systems where critical business data lives.
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
