The Compute Express Link™ (CXL) 3.0 IP Core is designed to meet the demands of next-generation heterogeneous computing systems, enabling high-speed, low-latency interconnect between CPUs, GPUs, FPGAs, DPUs, accelerators, memory expanders, and smart I/O devices.
Supporting the CXL 3.0 specification, this IP delivers backward compatibility with CXL 1.x/2.0, advanced memory pooling, and fabric capabilities, making it ideal for data centers, AI/ML workloads, and HPC environments.