Surprisingly Useful AI Article Enhancements
Cisco has launched a new silicon platform and infrastructure stack aimed at helping enterprises and service providers scale AI workloads while managing sovereignty, network complexity and emerging agent-based threats.
The announcements were made at Cisco Live EMEA in Amsterdam, where more than 21,000 IT professionals gathered for the event.
The new lineup includes the G300 switch silicon for high-density AI clusters, expanded observability and network control tools under the AgenticOps banner, upgrades to Cisco’s AI Defense product, and sovereign support infrastructure across Europe.
G300 Silicon Powers AI Buildouts
Cisco’s new Silicon One G300 is designed for high-performance AI clusters used in training, inference and real-time agentic workloads.
The G300 incorporates what Cisco calls Intelligent Collective Networking, which the company claims can lift network utilisation by 33 per cent and reduce job completion times by 28 per cent when compared to non-optimised environments.
G300 will ship in new N9100 and 8000 series systems targeted at AI builders including hyperscalers, neocloud providers, sovereign cloud deployments, service providers and enterprise data centres.
Cisco said the new hardware aims to improve operational efficiency and reduce the friction traditionally associated with GPU-centric compute clusters.
The company also introduced Nexus One, a unified management platform that simplifies operations across on-premises and cloud-based infrastructure.
AgenticOps Aims To Simplify Cross-Domain Complexity
AgenticOps, Cisco’s operational approach for agentic AI workloads, will now span its portfolio of networking, observability and security tools.
The system leverages telemetry from Cisco Networking, Security Cloud Control, Nexus One and Splunk to provide context-aware control of environments supporting autonomous agent workloads.
“AI is moving fast and so are the demands on IT teams,” said Jeetu Patel, President and Chief Product Officer at Cisco. “We’re bringing together silicon, systems, observability and control so that our customers can deploy safely and at scale.”
The AgenticOps tooling includes new platform capabilities to automate routing, threat analysis, performance monitoring and workflow control across multi-domain environments.
Upgrades To AI Defense And SASE
Cisco has updated its AI Defense product, expanding protections for organisations deploying agent-based AI systems.
The upgrades include AI supply chain governance, runtime protections and policy enforcement aimed at mitigating risks from manipulated or compromised agents.
Alongside this, the company has rolled out enhancements to its Secure Access Service Edge (SASE) offering. The update introduces AI-driven, intent-aware inspection of agentic traffic to assess the reasons behind tool invocations and requests.
Related: Best Business Laptops for work & school
Related: Best Gaming Laptops
Related: Best Portable Laptop
Cisco said the approach supports safe deployment of autonomous systems by enabling customers to monitor and filter agent interactions while maintaining the reliability of automated workflows.
Sovereignty Support And European CNSCs
Cisco has expanded its sovereign support model with services and infrastructure tailored to organisations that require strict data controls.
Its Customer Experience (CX) unit now offers support options for air-gapped, on-premises and hybrid environments, with controls aligned to jurisdictional requirements.
The company has also built out its Critical National Services Centre (CNSC) footprint across Europe, opening centres in the UK, France and Spain. Each operates under strict data-handling regimes with segregated processes, dedicated facilities and cleared personnel.
The CNSCs operate alongside Cisco’s regular support channels but offer secure paths for handling technical issues in highly regulated environments.
Cisco has operated a CNSC in Germany for over 15 years and said a fifth facility is now under development in Italy.
“These centres help us meet customers where they are, in compliance with how they want their data handled,” Patel said.
Last Updated on February 10, 2026 by Nick Ross
