Introduction to Compute Express Link (CXL) 3.0 - Features and Fabric Management

Learn about the latest advancements in the Compute Express Link™ (CXL™) 3.0 specification in this 35-minute technical presentation by AMD's Mahesh Wagh. Explore how modern cloud datacenters are evolving to meet the demands of compute-intensive workloads like AI and Machine Learning through fabric-based heterogeneous composable architecture. Discover the key improvements in CXL 3.0, including doubled data rates of 64GTs with no added latency, enhanced fabric capabilities, improved memory sharing and pooling, coherency, and peer-to-peer communication. Gain detailed insights into the CXL fabric management framework that enables fully composable and disaggregated memory for next-generation datacenter architectures, while maintaining backward compatibility with previous generations. Understand the implementation of multi-headed and fabric attached devices, enhanced fabric management, and composable disaggregated infrastructure that drive improved scalability and resource utilization in modern datacenter environments.