This episode will be available to stream at 9am ET.

As agentic AI systems move from experimentation to deployment, infrastructure must evolve to support real-time decision making, continuous interaction, and highly dynamic workloads. Building ultra-fast systems for these environments requires rethinking performance across the entire stack, from low-latency networking and high-speed interconnects to optimised compute and memory architectures. This session explores how organisations can design systems that meet the responsiveness and scale demands of agentic AI, while maintaining efficiency and reliability in distributed environments.

• Performance requirements for real-time and interactive AI systems
• Low-latency networking and high-speed interconnect strategies
• Optimising compute and memory for dynamic AI workloads
• Scaling infrastructure to support continuous, autonomous operations