3 Sources
[1]
Supermicro's datacenter-scale liquid cooling solution sets the stage for Nvidia Blackwell
Supermicro's new liquid cooling solution enables up to 96 B200 GPUs per rack. As GPUs for AI and HPC applications get more power-hungry, they require better cooling. For example, a rack with 72 Nvidia B200 GPUs is projected to consume around 120 kW, which is 10 times higher than a typical rack
[2]
Supermicro, Inc. Announces Complete Liquid Cooling Solution
Supermicro, Inc. announced a complete liquid cooling solution that includes powerful Coolant Distribution Units (CDUs), cold plates, Coolant Distribution Manifolds (CDMs), cooling towers and end to end management software. This complete solution reduces ongoing power costs and Day 0 hardware
[3]
Supermicro introduces full-scale liquid cooling for data centers By Investing.com
SAN JOSE, Calif. - Supermicro, Inc. (NASDAQ: SMCI), recognized for IT solutions for cloud and AI, has announced the launch of a comprehensive liquid cooling solution designed to reduce total cost of ownership (TCO) and enhance performance for large AI clusters. This solution, aimed at addressing
Share
Copy Link
Supermicro introduces a complete liquid cooling solution for data centers, designed to handle the extreme heat generated by next-generation AI and HPC servers, promising significant energy savings and improved performance.

Supermicro, a leader in IT solutions for cloud and AI, has unveiled a comprehensive liquid cooling solution designed to address the growing cooling demands of high-performance AI and HPC data centers
1
. This innovative technology aims to significantly reduce power consumption and improve overall data center efficiency.The new solution includes several advanced components:
This integrated system is capable of cooling servers that demand up to 12kW of power and AI racks generating over 100kW of heat
1
. The solution supports warm water cooling up to 113°F (45°C), allowing for potential heat reuse in applications such as district heating or greenhouse warming1
2
.Supermicro's liquid cooling technology enables ultra-dense AI servers, each housing dual top-tier CPUs and up to eight Nvidia HGX GPUs in a 4U configuration. This setup allows for up to 96 B200 GPUs per rack, quadrupling the computing density compared to traditional setups
1
2
.The system promises significant benefits:
2
3
One of the key advantages of Supermicro's solution is its rapid deployment capability. The modular design reduces data center setup time from months to weeks, crucial for companies looking to quickly upgrade or build new facilities
1
.The included SuperCloud Composer software offers real-time monitoring and control over all components, helping data center operators manage costs and maintain system reliability
1
2
.Supermicro has already implemented over 100,000 GPUs using this liquid cooling solution in some of the largest AI facilities ever built
1
3
. The company's CEO, Charles Liang, emphasized that their complete liquid cooling solutions are now cooling massive, state-of-the-art AI factories, reducing costs and improving performance1
.Related Stories
Supermicro's innovative cooling solution aligns with its strong financial performance. The company reported record annual revenues of $14.94 billion and a 109.77% revenue increase in the last twelve months as of Q4 2024
3
. Analysts anticipate continued sales growth, reflecting the expanding demand for energy-efficient data center solutions3
.As AI and HPC applications continue to demand more powerful and energy-intensive hardware, Supermicro's liquid cooling technology positions the company at the forefront of sustainable data center solutions. The ability to efficiently cool next-generation GPUs like Nvidia's Blackwell series sets the stage for future advancements in AI computing infrastructure
1
2
3
.Summarized by
Navi
[1]
[2]
1
Science and Research

2
Technology
3
Policy and Regulation
