F5 and NVIDIA: Innovations Unveiled at AppWorld Singapore 2025

Key Concepts:

• AI Innovation
• LLM (Large Language Model) Routing
• Inference Microservices (Dynamo)
• Token Counting
• MCP (Management Control Plane) Security
• NVIDIA Partnership
• Multi-tenancy
• Data Center Optimization
• AI Factories

AppWorld 2025 Singapore & AI Innovation

• Buu (Director of Community Evangelism at F5) interviews Ahmed Guetari (leading the Service Provider business for F5) at AppWorld 2025 in Singapore.
• Ahmed describes AppWorld Singapore as an "energizing event" with significant innovation and customer engagement, highlighting numerous projects and real-world use cases related to AI.
• Ahmed emphasizes the rapid pace of AI innovation, stating that "AI is extremely fast, right? You need to get on that train. It's a bullet train. And if you don't innovate fast, you're not going to keep up with it."

F5 & NVIDIA Partnership: Performance, Multi-tenancy, and Security

• F5 announced a partnership with NVIDIA in March at GTC, focusing on performance, multi-tenancy, and security.
• New innovations building upon this partnership are being developed in close collaboration with customers.

LLM Routing and Inference Microservices

• F5 is introducing LLM routing in partnership with NVIDIA, utilizing inferencing microservices.
• This allows customers to optimize their infrastructure by selecting the appropriate LLM for each prompt, without performance degradation.
• Integration with NVIDIA's Inference Microservices (Dynamo) optimizes inferencing, dramatically improving the number of prompts and the performance of the inference for the overall cluster.

Token Counting Capabilities

• F5 is introducing token counting capabilities, enabling precise control and limitation of token usage by organizations, groups, or teams.

MCP Security

• F5 provides security for MCP (Management Control Plane) services, addressing the security challenges associated with MCP deployments.
• F5 sits in front of the MCP, providing authentication, authorization, and Layer 7 protection for all traffic to the MCP server.

Future Innovations

• Ahmed hints at further innovations to be announced later in the year, emphasizing the continuous and rapid pace of development.

Data Center Optimization and AI Factories

• The innovations are crucial for optimizing data centers, especially in regions like Singapore and Malaysia, where significant data center construction is underway.
• These technologies are particularly relevant for "AI factories," which require high-performance and optimized infrastructure.

Conclusion

The conversation highlights F5's commitment to AI innovation, particularly through its partnership with NVIDIA. The focus is on optimizing performance, security, and multi-tenancy for AI workloads, with specific solutions like LLM routing, inference microservice optimization, token counting, and MCP security. These advancements are crucial for enabling efficient and secure AI deployments in modern data centers and AI factories.