At its annual global innovation event, Tech World, Lenovo (HKSE: 992) (ADR: LNVGY) introduced the ThinkSystem N1380 Neptune, a groundbreaking 6th generation vertical liquid cooling solution designed to optimize generative AI computing. This innovation enables organizations to efficiently run accelerated AI workloads while potentially reducing data center power consumption by up to 40%. Lenovo also announced an expanded partnership with NVIDIA, unveiling the ThinkSystem SC777 V4 Neptune, which integrates the N1380 Neptune technology to support the NVIDIA Blackwell platform and the NVIDIA GB200, facilitating the deployment of trillion-parameter AI models in a compact system.
As the demand for AI surges, data centers must adapt to higher thermal densities. The ThinkSystem N1380 Neptune and SC777 V4 Neptune are paving the way for data centers that can support 100KW+ server racks without specialized air conditioning. This advancement is crucial for modern data center designs that prioritize dense configurations and power efficiency.
Lenovo’s Chairman and CEO, Yuanqing Yang, stated, “For over a decade, Lenovo has been at the forefront of liquid cooling innovations, aiming to bring high-performance computing capabilities to all organizations. The ThinkSystem N1380 Neptune is set to transform data center efficiency and design, making NVIDIA Blackwell and trillion-parameter AI accessible while redefining power usage in data centers.”
The ThinkSystem N1380 Neptune maximizes power for AI and research, rather than for cooling systems. Its vertical liquid cooling chassis ensures accelerated computing with high efficiency in a fully liquid-cooled design, removing the need for energy-intensive fans. Compatible with standard 19-inch racks, this technology allows organizations to harness top-tier accelerated computing, leveraging standard power in an open ecosystem.
NVIDIA’s Bob Pette emphasized the significance of this collaboration, stating, “The NVIDIA Blackwell platform powers generative AI and heralds a new industrial revolution. With the Lenovo ThinkSystem N1380 Neptune and NVIDIA GB200, we’re poised to support the world’s most advanced accelerated workloads while significantly cutting operational costs and energy use, bringing AI’s potential to every industry.”
The ThinkSystem SC777 V4 Neptune also supports the next-generation NVIDIA Quantum-X800 InfiniBand and Spectrum-X800 Ethernet platforms for enhanced networking capabilities. Additionally, it includes NVIDIA AI Enterprise, a cloud-native software platform that simplifies the development and deployment of production-ready AI solutions across various domains, including generative AI and computer vision.
NVIDIA GB200: Redefining Enterprise AI
Lenovo will deliver NVIDIA GB200 rack systems that enhance AI training, data processing, and simulation. Featuring direct warm-water cooling for the entire NVIDIA GB200 system architecture, enterprise customers can run trillion-parameter AI models, accelerating product development while minimizing costs and maximizing energy efficiency.
The advanced cooling technology in the ThinkSystem server allows crucial components to operate at lower temperatures, effectively removing heat from GPUs, memory, and other components. This innovation enables AI training and real-time LLM inference for models scaling up to 10 trillion parameters, enhancing performance, energy efficiency, and reliability in a compact, ultra-quiet system.
Lenovo Neptune: A Leader in Liquid Cooling Technology
Lenovo continues to lead the liquid cooling market, leveraging over a decade of expertise and 40+ patents in the field. The Neptune platform utilizes state-of-the-art materials, including custom copper water loops and patented CPU cold plates, ensuring effective full-system cooling. Unlike traditional systems that use low-quality materials, Neptune employs durable stainless steel and reliable EPDM hoses for longevity and efficiency.
Supporting AI Deployment at Scale
With the launch of the 6th generation Lenovo Neptune, businesses can leverage the full hybrid AI portfolio, enabling rapid AI deployment and scalability. Lenovo TruScale GPUaaS offers on-demand access to GPU resources, simplifying management and optimization of AI workloads. This service will support a wide range of NVIDIA Tensor Core GPU options, accelerating access to NVIDIA’s accelerated computing for swift real-world innovation.
Lenovo’s Tech World ‘24 was live-streamed globally, allowing attendees to explore its hybrid AI solutions and learn more about the ThinkSystem N1380 Neptune. For further details, visit Lenovo Tech World ‘24.
Trademarks: LENOVO, THINKSYSTEM, and NEPTUNE are trademarks of Lenovo. NVIDIA is a trademark of NVIDIA Corporation, Inc. All other trademarks are the property of their respective owners. ©2024 Lenovo Group Limited.
