AI, DPUs, and Distributed Workloads: What F5 and Nutanix Are Building Together

Key Concepts

• Flow: Nutanix’s networking and network security product portfolio.
• MCN (Multi-Cloud Networking): Networking architecture that connects workloads across on-premises and multiple public clouds (AWS, Azure, GCP).
• DPU (Data Processing Unit): Specialized hardware used to offload networking and security tasks from the main CPU.
• Ingress Controller: A component that manages external access to services in a Kubernetes cluster.
• Digital Sovereignty: The requirement for data and workloads to reside within specific geographic or regulatory boundaries.
• Inference: The phase of AI where trained models are used to make predictions or process data.

1. Nutanix and F5 Strategic Partnership

The partnership between Nutanix and F5 focuses on creating a consistent "fabric" for application workloads, allowing them to run anywhere while eliminating discrepancies between different public cloud environments. Key integration points include:

• Application Load Balancing: F5 BIG-IP provides LTM (Local Traffic Manager) capabilities for workloads running on the Nutanix platform, serving as a primary migration path for customers moving away from VMware.
• Multi-Cloud Networking (MCN): F5 Distributed Cloud (F5XC) is utilized to provide seamless connectivity across Nutanix environments, whether on-premises or in public clouds (AWS, Azure, GCP).
• Kubernetes Ingress: NGINX Plus serves as the ingress solution for Nutanix’s Kubernetes platform. The companies are also collaborating on the transition from legacy ingress to modern Gateway APIs.

2. AI Infrastructure and Hardware Offloading

Nutanix is positioning itself as a platform for AI, specifically focusing on the transition from model training to inference.

• Hardware Ecosystem: Nutanix is collaborating with Nvidia and AMD to build prescriptive hardware/software stacks for AI workloads.
• DPU Integration: A significant technical focus is offloading networking and security stacks from the compute host to Nvidia or AMD DPUs. This frees up valuable CPU resources for AI workloads.
• Distributed AI: Due to the scarcity of GPUs, Nutanix and F5 are enabling "distributed AI" architectures. This allows enterprises to run workloads in one location while accessing scarce GPU resources located elsewhere, effectively bridging the gap between where data lives and where compute power is available.

3. Addressing Digital Sovereignty

Digital sovereignty has become a primary concern for Nutanix customers. The company is actively developing solutions to ensure that cloud offerings meet regional regulatory requirements, allowing organizations to maintain control over where their data and applications reside.

4. Future Outlook and Events

• Upcoming Announcements: The speakers hinted at "forward-looking" joint solutions currently in development, with more details to be shared at upcoming industry events.
• Nutanix .NEXT: The next major event is scheduled for April 7th–9th in Chicago, which serves as the Nutanix equivalent to F5 AppWorld.

5. Notable Quotes

• “Nutanix is a full stack alternative to VMware... together the combination of Nutanix and F5 provides this consistent fabric for application workloads to run any workload anywhere.” — Ani Duptari, Senior Director of Product, Nutanix.
• “Given the market conditions and supply chain constraints... GPUs are going to be a finite or a scarce resource... we are enabling this distributed AI workload... across the combination of Nutanix offerings and F5 offerings.” — Ani Duptari.

6. Resources for Further Information

• Main Portal: nutanix.com
• Joint Solutions: nutanix.com/partner (Navigate to the F5 section for specific integration details).

Synthesis

The collaboration between Nutanix and F5 is centered on simplifying the complexity of hybrid and multi-cloud environments. By integrating F5’s load balancing and MCN capabilities with Nutanix’s infrastructure, the partnership provides a unified platform that supports both traditional enterprise applications and modern, resource-intensive AI workloads. The strategic use of DPU offloading and the focus on digital sovereignty highlight a shift toward more efficient, hardware-accelerated, and compliant cloud architectures.