End to End Management of a CXL Enabled Datacenter Memory Fabric

Learn how to implement and manage Compute Express Link (CXL) technology for creating dynamic, scalable datacenter memory architectures in this 32-minute conference talk from the Open Compute Project. Explore the fundamental shift from traditional fixed memory configurations to fabric-based memory systems that decouple memory from compute resources, enabling unprecedented flexibility and efficiency in data-intensive applications. Discover the key architectural components including CXL-attached memory modules, switching infrastructure, and composable software orchestration systems that form the foundation of modern datacenter memory fabrics. Examine real-world deployment strategies and best practices for integrating advanced CXL memory technologies with high-performance CXL switches and fabric management software. Understand how to create fully dynamic, software-defined memory pools that can be provisioned, scaled, and managed in real time to meet the evolving demands of modern datacenter workloads. Gain insights from industry experts representing leading companies in CXL technology development, including perspectives on memory innovation, platform architecture, and enterprise-grade implementation strategies for transforming traditional server architectures into flexible, scalable computing environments.