SAN JOSE, CA, Aug 13, 2025 – Supermicro has expanded its NVIDIA Blackwell system portfolio with a 4U DLC-2 liquid-cooled NVIDIA HGX B200 system, available for volume shipment, and an air-cooled 8U front I/O system. The systems are designed to support NVIDIA HGX B300 platforms and provide front I/O access, simplified cabling, improved thermal performance, increased compute density, and lower operational expenses (OPEX).
“Supermicro’s DLC-2 enabled NVIDIA HGX B200 system leads our portfolio to achieve greater power savings and faster time to online for AI Factory deployments,” said Charles Liang, CEO and president, Supermicro. “Our Building Block architecture enables us to quickly deliver solutions exactly as our customers request. Supermicro’s extensive portfolio now can offer precisely optimized NVIDIA Blackwell solutions to a diverse range of AI infrastructure environments, whether deploying into an air- or liquid-cooled facility.”
Supermicro’s DLC-2 is a direct liquid cooling solution for AI-focused data centers. It provides cooling for high-density computing systems. The design supports operational efficiency and cost management in large-scale computing environments.
- Up to 40% data center power savings
- Faster time-to-deployment and reduced time-to-online by providing an end-to-end data center scale liquid cooling solution
- Up to 40% reduced water consumption with warm water cooling at an inlet temperature of up to 45°C, reducing the necessity of chillers
- Up to 98% system heat capture by liquid-cooling CPUs, GPUs, DIMMs, PCIe switches, VRMs, power supplies, and more
- Enabling quiet data center operation at a noise level as low as 50dB
Supermicro provides a portfolio of NVIDIA HGX B200 solutions, including two front I/O systems and six rear I/O systems. These options let customers select CPUs, memory, networking, storage, and cooling configurations that fit their needs. The new 4U and 8U front I/O NVIDIA HGX B200 systems are designed for large-scale AI training and inference. They address deployment issues like networking, cabling, and thermals.
“Advanced infrastructure is accelerating the AI industrial revolution for every industry,” said Kaustubh Sanghani, vice president of GPU product management at NVIDIA. “Based on the latest NVIDIA Blackwell architecture, Supermicro’s new front I/O B200 systems equip enterprises to deploy and scale AI at unprecedented speed – delivering breakthrough innovation, efficiency, and operational excellence.”
Modern AI data centers demand high scalability that requires substantial node-to-node connections. The system’s 8 high-performance 400G NVIDIA ConnectX-7 NICs and 2 NVIDIA Bluefield-3 DPUs are moved to the front of the system to allow for the configuration of networking cables, storage drive bays, and management all from the cold aisle. The NVIDIA Quantum-2 InfiniBand and the Spectrum-X Ethernet platform are supported to ensure the highest-performing compute fabric.
Supermicro has adjusted its components alongside changes to system architecture to improve efficiency, performance, and cost savings for AI data center tasks. The systems support memory expansion with 32 DIMM slots, giving users options for configuring memory and supporting larger memory capacities. This system memory works with the NVIDIA HGX B200’s HBM3e GPU memory to eliminate CPU-GPU bottlenecks, support processing of workloads, improve multi-job efficiency in virtualized settings, and accelerate data preprocessing.
All Supermicro 4U liquid-cooled systems and 8U or 10U air-cooled systems are optimized for NVIDIA HGX B200 8-GPU with each GPU connected via 5th Generation NVLink at 1.8TB/s, providing a combined total of 1.4TB of HBM3e GPU memory per system. NVIDIA’s Blackwell platform delivers up to 15x faster real-time inference performance and 3x faster training for LLMs compared to the Hopper generation of GPUs. The front I/O systems with dual-socket CPUs supporting up to 350W Intel Xeon 6 6700 series processors deliver high performance for a range of AI workloads.
The 4U front I/O liquid-cooled system features front-accessible NICs, DPUs, storage, and management components. It utilizes dual-socket Intel Xeon 6700 series processors with P-cores up to 350W and NVIDIA HGX B200 8-GPU configuration (180GB HBM3e per GPU). The system supports 32 DIMMs with up to 8TB capacity at 5200MT/s or up to 4TB at 6400MT/s DDR5 RDIMM, plus 8 hot-swap E1.S NVMe storage drive bays and 2 M.2 NVMe boot drives. Network connectivity includes 8 single-port NVIDIA ConnectX-7 NICs or NVIDIA BlueField-3 SuperNICs and two dual-port NVIDIA BlueField-3 DPUs.
Supermicro developed the liquid-cooled system as a base for AI clusters that can scale to thousands of nodes. This design can reduce data center power use by up to 40% compared to air-cooling.
The 8U front I/O air-cooled system shares the same front-access architecture and core features but is intended for AI setups without liquid cooling infrastructure. It has a compact 8U form factor (compared to Supermicro’s 10U system), a CPU tray, and keeps the 6U-height GPU tray to support air cooling.
Source: Super Micro
About Super Micro Computer
Super Micro Computer Inc., or Supermicro, is a provider of high-performance server technology and green computing solutions. Founded in 1993 by Charles Liang and Sara Liu, the company is headquartered in San Jose, CA. Supermicro offers a complete range of products, including servers, storage systems, networking devices, and server management software, serving industries like enterprise data centers, cloud computing, artificial intelligence, 5G, and edge computing. As of June 2023, the company employs approximately 5,126 individuals globally. In the fiscal year 2024, Supermicro reported revenues of approximately $15 billion, reflecting growth driven by its innovative solutions and expanding market presence.