• Home
  • Events Calendar
  • Blueprint Guidelines
  • Privacy Policy
  • Subscribe to Daily Newsletter
  • NextGenInfra.io
No Result
View All Result
Converge Digest
Friday, May 16, 2025
  • Home
  • Events Calendar
  • Blueprint Guidelines
  • Privacy Policy
  • Subscribe to Daily Newsletter
  • NextGenInfra.io
No Result
View All Result
Converge Digest
No Result
View All Result

Home » IBM Unveils z17 Mainframe with On-Chip AI and Spyre Accelerator

IBM Unveils z17 Mainframe with On-Chip AI and Spyre Accelerator

April 8, 2025
in Semiconductors
A A
IBM Spyre™ Accelerator brings generative AI capabilities to IBM z17. (PRNewsfoto/IBM)

IBM Spyre™ Accelerator brings generative AI capabilities to IBM z17. (PRNewsfoto/IBM)

IBM unveiled the z17 mainframe, the latest evolution of its enterprise computing platform, designed to bring AI acceleration directly into core infrastructure. Central to the system is the new IBM Telum II processor, which features an on-chip AI accelerator capable of executing over 450 billion inference operations daily with sub-millisecond response times. The chip includes a 40% increase in cache size over its predecessor and enhanced compute capacity, expanding the z17’s performance envelope for transactional and real-time AI workloads.

Complementing Telum II, IBM introduced the Spyre Accelerator, a PCIe-based AI engine expected in Q4 2025. The Spyre Accelerator brings generative AI capabilities to the mainframe, optimized for multi-model inference and direct integration with enterprise data. Together, Telum II and Spyre support native execution of assistants and agents like watsonx Code Assistant and watsonx Assistant, providing enterprise users with AI-enhanced development tools and real-time incident management through integrations such as Z Operations Unite.

IBM Telum® II Chip & IBM Spyre Accelerator Chip for AI Acceleration

IBM z17 will launch with support for the upcoming z/OS 3.2 operating system, which adds hardware-accelerated AI support, NoSQL compatibility, and hybrid cloud data processing. Enhancements also include new OpenTelemetry-based observability tools and expanded AI-driven cybersecurity features. IBM is also deploying Vault (from HashiCorp) for integrated secrets management, and Threat Detection for z/OS, a Telum-powered AI anomaly detection system designed to harden mission-critical workloads.

• Processor architecture:

• IBM Telum II with integrated AI acceleration

• 40% larger cache, higher frequency, expanded compute

• Up to 450B AI inference ops/day; 1ms response times

• AI hardware acceleration:

• IBM Spyre Accelerator via PCIe (Q4 2025)

• Optimized for generative AI, assistants, and agents

• Supports Granite-based models, runs enterprise AI workloads natively

• AI and OS software integration:

• z/OS 3.2 (expected Q3 2025) adds AI ops insight, NoSQL, hybrid cloud support

• watsonx Code Assistant and watsonx Assistant for Z integrated with Z Operations Unite

• Unified monitoring and response via IBM Concert and OpenTelemetry

• Security and observability:

• IBM Vault (by HashiCorp) for hybrid secrets management

• AI-based sensitive data discovery and classification

• IBM Threat Detection for z/OS for anomaly identification

• Storage and availability:

• IBM Storage DS8000 Gen 10 optimized for z17

• z17 general availability: June 18, 2025

• Spyre Accelerator availability: Q4 2025

“With z17, we’re bringing AI to the core of the enterprise with the software, processing power, and storage to make AI operational quickly,” said Ross Mauri, General Manager of IBM Z and LinuxONE.


Tags: IBM
ShareTweetShare
Previous Post

O-RAN Software Community Launches J and K Releases to Advance Open RAN

Next Post

UALink 1.0 Released for Low-Latency Scale-Up AI Accelerators

Related Posts

Lumen Brings AI to the Edge with IBM watsonx and Sub-5ms Infrastructure
Clouds and Carriers

Lumen Brings AI to the Edge with IBM watsonx and Sub-5ms Infrastructure

May 6, 2025
Netskope Integrates with OpenAI’s ChatGPT for Enhanced Security
Enterprise

Palo Alto Networks and IBM to Secure UK’s Emergency Services Net

January 15, 2025
IBM Develops Co-Packaged Optics Integrating PWG
Optical

IBM Develops Co-Packaged Optics Integrating PWG

December 10, 2024
New Plans for 324 MW Hyperscale Campus in Atlanta
All

IBM Rolls Out Guardium Data Security Center for Hybrid Cloud and Quantum Protection

October 22, 2024
IBM Expands Quantum Data Center in New York
Quantum

IBM Expands Quantum Data Center in New York

September 29, 2024
Intel Gaudi 3 AI Lands on IBM Cloud in 2025
Semiconductors

Intel Gaudi 3 AI Lands on IBM Cloud in 2025

August 29, 2024
Next Post
UALink 1.0 Released for Low-Latency Scale-Up AI Accelerators

UALink 1.0 Released for Low-Latency Scale-Up AI Accelerators

Categories

  • 5G / 6G / Wi-Fi
  • All
  • Automotive Networking
  • Blueprints
  • Clouds and Carriers
  • Data Centers
  • Enterprise
  • Explainer
  • Financials
  • Last Mile / Middle Mile
  • Optical
  • Quantum
  • Research
  • Security
  • Semiconductors
  • Space
  • Start-ups
  • Subsea
  • Sustainability
  • Video
  • Webinars

Archives

Tags

5G All AT&T AWS Blueprint columns BroadbandWireless Broadcom China Ciena Cisco Data Centers Dell'Oro Ericsson FCC Financial Financials Huawei Infinera Intel Japan Juniper Last Mile Last Mille LTE Mergers and Acquisitions Mobile NFV Nokia Optical Packet Systems PacketVoice People Regulatory Satellite SDN Security Service Providers Silicon Silicon Valley StandardsWatch Storage TTP UK Verizon Wi-Fi
Converge Digest

A private dossier for networking and telecoms

Follow Us

  • Home
  • Events Calendar
  • Blueprint Guidelines
  • Privacy Policy
  • Subscribe to Daily Newsletter
  • NextGenInfra.io

© 2025 Converge Digest - A private dossier for networking and telecoms.

No Result
View All Result
  • Home
  • Events Calendar
  • Blueprint Guidelines
  • Privacy Policy
  • Subscribe to Daily Newsletter
  • NextGenInfra.io

© 2025 Converge Digest - A private dossier for networking and telecoms.

This website uses cookies. By continuing to use this website you are giving consent to cookies being used. Visit our Privacy and Cookie Policy.
Go to mobile version