Enterprise AI is transferring quickly from piloting to manufacturing. As organizations deploy bigger fashions and agentic AI methods, they shortly notice that AI is simply as quick—and as safe—because the community it runs on. AI workloads demand large bandwidth, low latency, and sturdy safety that conventional networks battle to offer.
To bridge this hole, infrastructure groups want options that ship unprecedented scale, operational simplicity, and pervasive safety. We’re excited to showcase precisely how we’re making this a actuality on the ONUG AI Networking Summit Dallas, going down Might 13–14, 2026, in Frisco, Texas.
Here’s a take a look at what you may anticipate from Cisco on the occasion and the way we may help you construct your AI-native enterprise.
Catch Cisco’s thought leaders on the principle stage and past
The shift towards AI-ready infrastructure requires new frameworks and operational fashions. Our consultants will likely be sharing actionable insights throughout a number of key periods:
When Brokers Run the Community: Safety, Entry, and the AI Information Middle Forward
Primary Stage Keynote | Might 13
The rise of AI brokers modifications how networks are constructed and secured. Tom Gillis, Cisco SVP & GM of the Infrastructure and Safety Group, will ship a forward-looking keynote on the AI-native knowledge heart. He’ll discover how networks should adapt for machine-to-machine intelligence and what safety fashions are required when autonomous brokers drive operations.
From AI Ambition to AI Infrastructure: Your Blueprint to Success for Enterprise Networks
Cisco Triple-T Session | Might 13
Earlier than you may run AI effectively, you want a wise, safe community constructed to hold it. This session by Sai Natarajan, Sr. Director, Cisco Information Middle product administration, cuts via the complexity, providing a sensible blueprint for deploying AI infrastructure with out disrupting present workloads. Learn to simplify, future-proof, and supercharge your knowledge heart for the Agentic Period.
Hybrid AI Infrastructure and AgenticOps Safety in Follow
Panel Session | Might 14
Tom Gillis joins a panel of business leaders to debate constructing hybrid AI infrastructures that steadiness public AI providers with personal mannequin deployments. Uncover how aligned architectures and operational fashions create safe, adaptable environments for altering AI consumption patterns.
Defending In opposition to Adversarial AI Brokers – From Digital Darwinism to Guardrails
Panel Session | Might 13
As AI brokers develop into extra autonomous, they introduce new assault vectors. Craig Connors, VP and CTO of the Cisco Safety Enterprise Group, will be part of this panel to debate how adversarial brokers change risk modeling and what defensive patterns, guardrails, and knowledge controls are required to mitigate these rising dangers.
Improvements powering the AI enterprise
At ONUG, we’ll spotlight the newest Cisco improvements designed to scale out and scale throughout for the agentic period, guaranteeing your transition to manufacturing AI is seamless and safe.
Unprecedented scale and ease
AI workloads generate visitors depth that pushes typical networks to the breaking level. We’re answering this problem with Cisco Nexus One, powered by the Silicon, Techniques, Optics, Software program, Operational Fashions, Safety and Observability.
Moreover, we’re leveraging AgenticOps through AI Canvas, bringing AI-driven troubleshooting and optimization to your enterprise infrastructure utilizing the Cisco Deep Community Mannequin.
To make sure each AI job runs at peak effectivity, Cisco’s structure brings collectively superior AI job monitoring, clever blended mode load balancing, and built-in GPU optimization throughout the material. Actual-time monitoring supplies visibility into job standing and useful resource allocation, so infrastructure groups can determine and resolve congestion or bottlenecks earlier than they impression efficiency. Combined mode load balancing dynamically shifts workloads throughout community paths and clusters for optimum circulate, minimizing latency and maximizing throughput—even in periods of unpredictable demand. Mixed with deep integration into GPU orchestration, this permits organizations to spice up compute utilization, velocity up AI mannequin convergence, and finally drive larger enterprise worth from each GPU funding.
The outcome: your AI workloads are executed extra effectively, reliably, and with agility, empowering you to maneuver sooner from proof-of-concept to impactful manufacturing deployments.
Deep, pervasive safety for AI workloads
You can’t scale AI with out securing it. Conventional safety fashions usually create bottlenecks or tax the very CPUs wanted for AI processing. We’re altering the sport with a unified safety method embedded immediately into the community.
Our Cisco N9300 Sequence Sensible Switches fuse networking and safety, supporting L4 segmentation to make sure insurance policies observe the workload. Mixed with Cisco Hypershield, organizations acquire air-gapped, distributed segmentation throughout the material. To guard towards vulnerabilities in actual time, Cisco Dwell Shield immediately deploys eBPF-based shields with out requiring reboots or downtime.
For the last word in server-level safety, we’re showcasing our integration of Cisco Hybrid Mesh Firewall with NVIDIA BlueField DPUs. By operating stateful segmentation and firewall insurance policies immediately on the DPU, we ship line-rate safety with out compromising GPU or CPU efficiency. This ensures your AI front-end materials stay safe, remoted, and extremely environment friendly.
Let’s construct the way forward for AI collectively
The AI revolution is right here, and the community is its basis. Whether or not it is advisable scale large GPU clusters, safe multi-tenant AI environments, or simplify operations with agentic automation, Cisco has the blueprint to get you there.
We look ahead to being part of the ONUG neighborhood and seeing you in Dallas. Be sure you attend our periods and join with our consultants to debate your AI networking and safety methods.
