View the CXL video demos to explore the benefits of CXL technology for AI applications.
Astera Labs’ Expanded Leo Smart Memory Controller™ Family
See Astera Labs new Leo X-Series for fabric-attached KV cache offload and the Leo 2 E-Series and P-Series for CXL 3.2 expansion, pooling, and sharing.
Montage Technology: CXL Memory Expansion: Real-World Performance Demo Using Montage’s CXL MXC Controller
In this demo, Montage compares a traditional 1TB DRAM-only server to a hybrid 1TB system with 512GB DDR5 DRAM and 512GB CXL memory provided by four RDIMMs through Montage’s MXC add-in cards. Configured in flat memory mode, the system runs MongoDB and YCSB across sockets to simulate real-world workloads. Despite using CXL-attached memory, the system achieves 95–100% of the throughput delivered by DRAM alone, with only 5–10µs added application-level latency. Montage’s CXL MXC controller contributes just 70ns of delay and provides a scalable, production-ready, and cost-efficient path to expand memory for data-intensive environments.
Netlist: NVvault Demonstration
Netlist’s NVvault™ demonstration showcases a CXL Type 3 persistent memory solution that combines DRAM, NAND flash backup, and onboard backup power to deliver memory-class performance with data persistence across both controlled shutdowns and unexpected power loss events. The demo highlights CXL.mem, cache-coherent host integration, Global Persistent Flush (GPF), automatic save/restore functionality, and persistent-memory use cases including databases, analytics, storage acceleration, and HPC checkpointing. Additional information is available at www.netlist.com and https://netlist.com/cxl-nvvault/.
Penguin Solutions: Enabling Next-Generation AI and HPC Infrastructure with SMART™ CXL Memory Expansion
As AI and HPC workloads continue to push the limits of traditional server memory architectures, organizations need scalable, cost-effective ways to increase memory capacity and utilization. Discover how Penguin Solutions’ SMART™ CXL Memory Expansion Add-In Cards (AICs) leverage Compute Express Link (CXL) technology to expand memory resources without major platform redesigns. By enabling greater memory scalability, improved resource efficiency, and enhanced system flexibility, SMART CXL solutions help AI, HPC, and data center operators address the growing demands of large-scale training, inference, simulation, and data-intensive computing workloads.
Primemas’ Disaggregated Memory Solutions: The Complete CXL Memory Stack
Primemas DDR Memory Solutions offers scalable, CXL-based DDR memory solutions from <1 TB to> 1 PB and beyond.The company’s CXL memory ecosystem includes PMA14 and PMA16 CXL Memory AICs, supporting direct-attached memory expansion, JBOM (Just a Bunch of Memory), and SLiM (Switchless Pooled Memory) configurations based on the CXL 3.x specification. The demo showcases low-latency memory expansion, switchless pooled memory in a compact 1U platform, and JBOM with terabyte-scale DDR5 capacity, and the Falcon SoC chiplet architecture for scalable memory and compute acceleration. Together, these technologies enable flexible, disaggregated memory solutions for AI, cloud infrastructure, and real-time analytics. Visit www.primemas.com for more information.