Managing Your AI Data Center at Scale with Juniper Networks

This presentation explores how Juniper Networks' Apstra solution effectively manages AI data centers at scale. Learn how Apstra simplifies network configuration for AI/ML workloads by providing tools to assign virtual networks across numerous ports—a critical capability in environments with potentially millions of ports. Discover the intent-based approach that makes provisioning virtual networks and configuring load balancing straightforward, reducing manual configuration burdens and enabling quick deployment regardless of GPU count. See how continuous validation features help catch configuration issues like missing VLAN assignments before they impact operations, while bulk operations streamline virtual network and subnet assignments across the entire infrastructure. The presentation demonstrates Juniper's focus on an intuitive interface that minimizes command-line configuration needs, making AI data center deployment faster and more efficient. Kyle Baxter also discusses Apstra's delivery as a virtual machine available for download, its on-premise deployment capabilities, integration with MIST AI for AI-driven network operations, and upcoming version 6.0 features including RDMA load balancing support. Presented by Kyle Baxter, Head of Apstra Product Management at Juniper Networks, recorded in Santa Clara, California on April 23, 2025, as part of AI Infrastructure Field Day.