
NVIDIA Blackwell helps achieve a 20% liquid cooling penetration rate by 2025, Lenovo is going to win big

According to TrendForce's survey, NVIDIA's Blackwell platform is expected to ship in the fourth quarter of 2024, driving the penetration rate of liquid cooling solutions from 10% in 2024 to 20% in 2025. With the increasing global ESG awareness and cloud service providers accelerating the construction of AI servers, the demand for liquid cooling solutions will significantly increase. NVIDIA currently holds nearly 90% of the market share in the AI server market, and it is expected that by 2025, the Blackwell platform will account for over 80% of the high-end GPU market
According to the latest survey by TrendForce, with the expected launch of the NVIDIA Blackwell new platform in the fourth quarter of 2024, it will drive a significant increase in the penetration rate of liquid cooling solutions, from around 10% in 2024 to over 20% in 2025. As global awareness of ESG (environmental, social, and corporate governance) rises, coupled with CSPs (cloud service providers) accelerating the construction of AI servers, it is expected to help shift cooling solutions from air-cooled to liquid-cooled forms.
In the global AI server market observation, in 2024, the main AI solution provider is still NVIDIA. In terms of GPU AI servers, NVIDIA holds a strong advantage with a market share of close to 90%, while the second-ranked AMD accounts for only about 8%.
TrendForce observes that the shipment volume of NVIDIA Blackwell this year is relatively small mainly because the supply chain is still undergoing final product testing and verification processes, with continuous optimization needed in areas such as high-speed transmission and cooling design. The new platform, due to its high energy consumption, especially the GB200 full-cabinet solution requiring better cooling efficiency, is expected to promote the penetration rate of liquid cooling solutions. TrendForce estimates that by 2025, the Blackwell platform's share in high-end GPUs is expected to exceed 80%, which will drive power supply manufacturers and the cooling industry to actively enter the AI liquid cooling market, forming a new industry competitive landscape.
In recent years, large American cloud service providers such as Google, AWS, and Microsoft have accelerated the deployment of AI servers, mainly using NVIDIA GPUs and self-developed ASICs. According to TrendForce, the thermal design power (TDP) of the NVIDIA GB200 NVL72 cabinet is as high as about 140kW, requiring a liquid cooling solution to address the cooling issue. Other architectures such as HGX and MGX in the Blackwell servers have lower density, with air cooling as the main solution.
In terms of self-developed AI ASICs by cloud service providers, Google's TPU not only uses air cooling but also actively deploys liquid cooling, making it the most proactive American company in adopting liquid cooling solutions. In mainland China, Alibaba is the most actively expanding liquid-cooled data centers.
Currently, in the global AI server liquid cooling solutions, Lenovo, HP, and Inspur each have their own technologies, with Lenovo's Neptune liquid cooling technology being the most mature and efficient. For over a decade, Lenovo's Neptune has been at the forefront of data center cooling technology, pioneering a unique method of using liquid cooling that can achieve up to a 40% reduction in power consumption compared to traditional air cooling systems. The Neptune solution uses a warm water circulation to cool the data center systems and keeps all server components cool, reducing the need for high-power system fans during data center operation.
With over 40 industry-first patents, Lenovo is taking liquid cooling technology to new heights. With the introduction of the sixth-generation Lenovo Neptune liquid cooling technology, Lenovo is expanding its industry-leading Neptune liquid cooling technology to the mainstream product portfolio of ThinkSystem V3 and V4, and introducing new design innovations • Expanding the wide application of Neptune technology: The mainstream ThinkSystem V3 and V4 products offer the option of direct open-loop water cooling for CPUs.
• Brand new industry-exclusive cold plate design: A patented cold plate design optimized for CPUs and accelerators, maximizing the heat dissipation efficiency of approximately 700W accelerators, with future designs supporting over 1000W accelerators.
• Innovative memory cooling technology for the ThinkSystem platform: Lenovo's patented memory cooling technology designed for high-density high-performance computing (HPC) systems will be integrated into specific ThinkSystem platforms, utilizing direct liquid cooling for DRAM, allowing customers to choose CPU and memory cooling options to effectively reduce heat and maximize performance.
The sixth-generation Neptune warm water cooling design will allow operation without the need for special data center air conditioning, and by allowing the reuse of facility hot water, waste heat is converted into value.
Through these innovations, Lenovo continues to strengthen its leadership position in high-performance computing and sustainable development, providing customers with more efficient and environmentally friendly cooling solutions to meet the growing demand for high-performance computing
