index

(2024-10-28 Shanghai) USI (SSE 601231), a global provider of electronic design and manufacturing services, in collaboration with its partners and Intel, has successfully developed the OCP DC-MHS 2U cloud server system, based on Intel’s latest Birch Stream server platform. This system offers a powerful, flexible solution for computing and storage in data centers, designed to meet the evolving demands of high-performance cloud infrastructure.

Photo: OCP DC-MHS 2U Diagram

The cloud server system, OCP DC-MHS 2U, features the DC-MHS M-FLW dual-CPU motherboard, integrating modular extended Input/Output connectors and modular peripheral interface connector. These components, along with modular peripheral and storage interface, enable efficient high-performance computing. The system supports up to 32 DDR5 memory modules and accommodates various PCIe Gen5 external cards, including up to three Double-Wide Graphics Processing Units, or six Full-Height Full-Length cards, plus one Full-Height Half-Length card, and one Low-Profile card, offering significant flexibility to meet diverse workload requirements.

The system’s front supports up to eight 2.5-inch Non-Volatile Memory Express (NVMe) drives and either eight or 16 E1.S high-speed storage devices. At the rear, the system is equipped with OCP NIC r3 high-speed network interface cards and supports the data center secure control module along with two additional E1.S storage devices, ensuring robust networking and storage capabilities.

For power management, the system offers support for both common redundant power supply and modular common redundant power supply, catering to a range of power needs. To meet future requirements for high-power CPUs (over 300W), USI provides liquid cooling solutions, addressing the increasing demand for higher computational power and energy efficiency while ensuring stable, reliable operation in data centers.

USI’s server development team emphasized that the new OCP DC-MHS 2U cloud server system is not only designed to enhance computing performance and flexibility but also to offer scalability for future workloads. One of the key application areas for this server system is artificial intelligence (AI), as it can handle large-scale AI computing requirements, providing a stable and efficient computing platform for data-intensive AI applications. Its modular architecture allows data centers to easily upgrade components as technology evolves, achieving cost-effective. The team also noted that the system’s support for liquid cooling solutions will play a key role in optimizing energy efficiency, especially as demand for higher computational power increases. By reducing operational costs and minimizing environmental impact, this advanced server solution underscores USI’s commitment to sustainability and the future of green data centers.