SK Hynix Completes Customer Validation of CXL 2.0-Based DDR5 Module, Advancing Data Center Memory Innovation

April 23, 2025 — SK Hynix has successfully completed customer validation for its 96GB CXL Memory Module (CMM), marking a significant milestone in the adoption of CXL 2.0-based DRAM technology. Designed for next-generation data centers, this DDR5-based solution offers substantial gains in memory capacity, bandwidth, and operational efficiency.

April 23, 2025 — SK Hynix has successfully completed customer validation for its 96GB CXL Memory Module (CMM), marking a significant milestone in the adoption of CXL 2.0-based DRAM technology. Designed for next-generation data centers, this DDR5-based solution offers substantial gains in memory capacity, bandwidth, and operational efficiency.

According to SK Hynix, the new 96GB CXL DDR5 module delivers 50% more capacity and 30% higher bandwidth than conventional DDR5 memory modules. It enables data processing speeds of up to 36GB/s, significantly reducing the Total Cost of Ownership (TCO) for data center customers.

The company is also conducting validation for a 128GB CXL DDR5 module, which is equipped with fifth-generation 10nm-class (1b) 32Gb DDR5 DRAM. This model is engineered for improved performance-per-watt, supporting customers with higher-density, energy-efficient memory solutions. SK Hynix aims to complete validation quickly to ensure market readiness and strengthen its CXL product lineup.

In parallel with hardware innovation, SK Hynix has actively contributed to the CXL software ecosystem. The company has developed its proprietary Heterogeneous Memory Software Development Kit (HMSDK), optimized for CXL memory environments. This software was successfully integrated into the Linux operating system in 2024, significantly boosting CXL-based system performance.

SK Hynix continues to pursue its "Optimal Innovation" strategy by offering scalable, flexible, and cost-effective memory solutions tailored for AI-driven workloads. Through advances in both hardware and software, the company aims to resolve the limitations of traditional systems—namely high cost and limited scalability—and deliver enhanced value to customers across AI, enterprise, and cloud computing markets.

Related Articles