F5 Boosts AI Delivery with BIG-IP Next and NVIDIA DPUs

F5 has introduced BIG-IP Next for Kubernetes, AI application delivery and security solution, designed to accelerate, secure, and optimize data traffic for AI infrastructures in large enterprises and service providers. By integrating NVIDIA’s high-performance BlueField-3 Data Processing Units (DPUs), F5 aims to boost the efficiency and scalability of AI workloads while ensuring robust security and resource optimization.

The collaboration with NVIDIA enables F5’s BIG-IP Next for Kubernetes to handle high-performance networking, traffic management, and security, reducing latency and increasing resource utilization. This helps organizations better manage the vast amounts of data required for AI models, delivering faster AI inferences and improving the overall customer experience.

Enhanced Performance for AI Workloads

BIG-IP Next for Kubernetes, combined with NVIDIA’s BlueField-3 DPUs, offers organizations centralized control over AI data traffic management, enhancing the utilization of data center resources. This results in faster, more responsive AI inference, benefiting industries reliant on AI for real-time data processing, such as telecommunications, 5G, and cloud-native environments.

The technology streamlines key AI processes, from inference to retrieval-augmented generation (RAG) and data management. The integration minimizes hardware demands, reduces energy consumption, and improves multi-tenancy, making it an ideal solution for both the network edge and core in AI-driven environments.

Adapting to Cloud-Native Infrastructure

The solution is purpose-built for Kubernetes environments, aligning with the growing demand for cloud-native network functions (CNFs). Service providers and enterprises can now transition seamlessly to cloud-native infrastructure, offloading data-intensive tasks to the NVIDIA DPUs. This frees up CPU resources for critical AI applications, ensuring high performance while optimizing costs.

Strategic AI Control and Security

BIG-IP Next for Kubernetes offers deep observability, ingress and egress control, and support for multiple protocols, enhancing security across AI networks. Customers can automate AI endpoint discovery, ensuring data integrity while addressing the growing security needs of AI models.

Kunal Anand, Chief Technology and AI Officer, F5: “This integration between F5’s robust application delivery and NVIDIA’s accelerated computing enhances observability, control, and performance for AI workloads, creating a powerful solution for organizations managing large-scale AI infrastructure.”

Ash Bhalgat, Sr. Director of AI Networking and Security Partnerships, NVIDIA: “F5’s collaboration with NVIDIA provides a scalable solution to deliver high-performance AI applications securely, powered by BlueField-3 DPUs to enhance cloud-scale efficiency.”

The solution is particularly valuable for telco service providers as they shift to cloud-native infrastructure, offloading intensive tasks to BlueField-3 DPUs, freeing CPU resources for critical applications. It also strengthens edge computing and core network deployments for technologies like 5G and future advancements toward 6G.