On February 24, 2026, Purple Hat introduced a brand new collaboration with NVIDIA to offer a co-engineered software program platform for AI factories. Constructing on this momentum, we’re happy to announce that Cisco Safe AI Manufacturing unit with NVIDIA now has an expanded selection of tooling for AI practitioners to speed up AI software supply. That is enabled by the assist for Purple Hat AI Manufacturing unit software program. This mix of NVIDIA AI Enterprise and Purple Hat AI Enterprise software program helps simplify the event of next-generation agentic workloads by offering a single, built-in AI manufacturing facility stack. Not like different options that deal with safety as an afterthought, Cisco Safe AI Manufacturing unit with NVIDIA offers a uniquely built-in, security-first basis. By fusing safety and observability instantly into the community and each layer of the stack, we be sure that production-ready innovation is not only scalable however inherently protected against the core to the sting.
Scaling these production-ready workloads from an preliminary pilot to a world enterprise rollout requires a basis of high-performing, resilient infrastructure. Cisco Safe AI Manufacturing unit with NVIDIA offers this important platform, guaranteeing that each part (from Cisco accelerated compute and networking to safety and observability, to the Purple Hat and NVIDIA AI software program) operates with the reliability and velocity crucial for production-grade AI. By offering this industrial-scale answer, we allow organizations to maneuver past experimentation and confidently scale their AI initiatives throughout the enterprise.
The muse: Strategic engineering alignment
So how do you rework from AI experimentation to changing into a genuinely “AI-fueled” enterprise? It begins together with your infrastructure. Cisco brings the high-performance networking material, accelerated compute nodes, and built-in safety structure you might want to deal with the huge information throughput of contemporary AI clusters. Our management stems from deep, strategic engineering alignment with Purple Hat and NVIDIA, guaranteeing that Cisco UCS and Nexus infrastructure work seamlessly with the Purple Hat AI Manufacturing unit with NVIDIA software program stack.
By supporting Purple Hat AI Manufacturing unit software program as an possibility, we now have expanded buyer selection of AI software program in our streamlined answer with a co-engineered software program platform for AI factories. Cisco Safe AI Manufacturing unit with NVIDIA infrastructure comes pre-validated to run the business’s main enterprise Linux and MLOps software program, permitting IT groups to deploy with confidence and AI practitioners to start out working instantly.
Establishing operational rigor for AI practitioners
This engineering alignment between Purple Hat and NVIDIA tackles probably the most persistent roadblock to enterprise AI adoption: manufacturing supply friction. For AI practitioners, this implies shifting past experimental silos and right into a unified, high-performance atmosphere. Purple Hat AI Enterprise and NVIDIA AI Enterprise present the software program instruments to handle the AI/ML lifecycle at scale. The brand new Purple Hat Enterprise Linux (RHEL) for NVIDIA gives a specialised, validated working basis, and the true worth is realized by Cisco, Purple Hat, and NVIDIA coming collectively to speed up enterprise AI with Cisco Safe AI Manufacturing unit with NVIDIA.
By upstreaming proprietary drivers and GPU operators instantly into the RHEL ecosystem, Purple Hat and NVIDIA have eradicated the necessity for guide remediation. For Cisco clients, this software-level integration finds its good dwelling on Cisco AI PODs. Our engineering ensures that these AI clusters may be deployed, ruled, and operated with the identical rigor, safety, and predictability as your core enterprise IT platforms.
Whenever you layer this co-engineered software program stack onto Cisco unified compute and high-performance networking material, the “infrastructure tax” disappears. With Cisco Intersight offering clever, cloud-based administration, IT groups can simplify operations throughout information heart and edge environments, guaranteeing constant efficiency no matter location. This enables information scientists to cease troubleshooting driver variations or interconnect bottlenecks and begin specializing in mannequin accuracy and enterprise influence. With the Cisco Safe AI Manufacturing unit with NVIDIA, we aren’t simply providing you with the instruments to construct AI; we’re offering the industrial-grade basis to run it with confidence.
Accelerating the trail to agentic AI
Whereas operational stability lays the groundwork, you additionally want to beat the complexity that usually stalls deployment velocity. In at present’s market, velocity is a aggressive necessity. Purple Hat simplifies software program acquisition and deployment by a brand new, streamlined licensing mannequin. By combining Purple Hat AI Enterprise with NVIDIA AI Enterprise (accessible by Cisco) in a unified AI Manufacturing unit software program platform, Purple Hat and NVIDIA have made high-end MLOps economically accessible. This unified procurement by Cisco additional simplifies the trail to manufacturing by offering a single level of contact for each your infrastructure and the AI software program stack.
What’s extra, Cisco clients acquire instantaneous entry to pre-configured fashions, together with the IBM Granite household and NVIDIA Nemotrondelivered as NVIDIA NIM microservices. With Purple Hat AI Manufacturing unit software program now supported by Cisco, AI practitioners can skip the difficult setup part and bounce straight to fine-tuning and inference, considerably lowering time to worth for next-generation bodily and agentic AI functions.
Unparalleled hybrid cloud flexibility
As you speed up these deployments, the place you run your workloads turns into a crucial strategic choice. Does your AI technique require the dimensions of the general public cloud, the safety of an on-premises information heart, or the low latency of the sting? With Cisco assist for Purple Hat AI Manufacturing unit software program, the reply is “all the above.”
Powered by Purple Hat OpenShift Container Platformthis answer offers a constant operational mannequin throughout core to the sting. Cisco clients can construct and practice fashions on-premises on AI PODs, which kind the infrastructure basis of Cisco Safe AI Manufacturing unit with NVIDIA, whereas sustaining the flexibleness to deploy these fashions wherever. This ensures you preserve full architectural management over your information and fashions, no matter the place the bodily {hardware} resides.
Safety and efficiency at scale
Sustaining this architectural flexibility throughout the hybrid cloud requires simultaneous deal with the non-negotiable necessities of enterprise-grade safety and efficiency.
At its core, safety can’t be an afterthought. Siloed instruments and fragmented enforcement go away gaps, exposing AI environments to threats like mannequin poisoning, immediate injection, and information exfiltration. Cisco Safe AI Manufacturing unit with NVIDIA is constructed in another way; safety is engineered into each layer of the stack constantly from core to edge. Cisco AI Protectionbuilt-in with NVIDIA AI, protects AI fashions and functions from improvement to runtime—guaranteeing each inference is trusted and safeguarded in opposition to immediate injection, dangerous output, and information privateness dangers. Cisco Isovalent delivers runtime safety for containerized AI workloads on Kubernetes, stopping lateral motion by identity-based, kernel-level enforcement utilizing eBPF know-how, with out impacting efficiency. This safety extends to the infrastructure, with constant coverage enforcement by Cisco Hybrid Mesh Firewall and Cisco Safety Cloud Managementwhereas Cisco Splunk Enterprise Safety offers built-in menace detection and response. That is safety engineered into the AI Manufacturing unit, not bolted on.
Complementing that is Splunk Observability Cloud’s Observability for AIwhich delivers unified visibility throughout the Safe AI Manufacturing unit with NVIDIA to observe efficiency, high quality, safety, and value. With AI Infrastructure Monitoring, groups acquire real-time insights into the well being, availability, and useful resource utilization (e.g., GPU, energy, community, nodes) of Cisco AI PODs, guaranteeing value attribution and optimization. AI Agent Monitoring tracks LLM and agentic functions for hallucinations, biases, and semantic high quality, whereas detecting AI privateness dangers by integration with Cisco AI Protection. These real-time insights empower proactive root trigger evaluation and fast difficulty decision to enhance the effectivity, reliability, and safety of the whole AI stack.
The highway forward
By combining the sturdy infrastructure from Cisco with the revolutionary AI software program stacks from Purple Hat and NVIDIA, we’re turning the promise of AI right into a mission-critical actuality for each enterprise. This allows you to speed up the supply of trusted agentic and bodily AI functions with scalable, high-performance, and safe infrastructure.
Contact your Cisco account consultant at present to schedule a method session on Cisco Safe AI Manufacturing unit with NVIDIA. Collectively, we are able to assess your infrastructure readiness and chart a path towards industrial-scale AI manufacturing.
