Notably, two giants within the sector have launched a historic collaboration that will bring about a revolution in artificial intelligence infrastructure implementation. The collaboration integrates knowledge and expertise gathered from several decades in electrical engineering with highly advanced liquid cooling solutions. Notably, this collaboration marks a historic event within an ever-growing focus on developing scalable next-generation computing solutions.
Strategic collaboration tackles rising demands for AI infrastructure
Siemens and nVent are working on developing a liquid cooling and power reference architecture. As AI computing workloads become more demanding and increasingly distributed, data center infrastructure designs have increasingly had to address and resolve competing demands on performance, efficiency, and scale with more intelligent and adaptive solutions. Siemens contributes expertise with industrial-grade electrical systems, and nVent brings its expertise with liquid cooling.
Current data centers experience rising rack-power densities, an increase in computation-intensive workloads, and a need for modularity. Reference architectures significantly enable data center operators with rapid deployment and interfacing, as well as offer frameworks for innovation. All these factors are met by the collaboration, as it addresses these needs with comprehensive engineering solutions for various data center implementation scenarios. Specifically, the project will offer considerable acceleration within strict energy efficiency and sustainability goals.
Modular blueprint facilitates faster deployment strategies
The combined reference architecture represents a modular, fault-tolerant, and energy-efficient solution that would enable operators to more swiftly and intelligently implement AI infrastructure compared with previous methods. The impact of complexity on overall operations would be eliminated with this streamlined solution, yet it would still be flexible enough for various operations. Sara Zawoyski, president of nVent Systems Protection, pointed out that nVent had decades of experience serving its customers’ next-generation computing infrastructure needs.
Advanced architecture targets hyperscale computing environments
A new joint architecture has been developed to enable customers to construct 100 MW hyperscale AI data centers for housing large-scale liquid-cooled AI infrastructures such as NVIDIA DGX GB200. It depicts a Tier III-ready infrastructure that unites Siemens’ industrial-grade electrical and automation solutions with NVIDIA DGX reference architectures and nVent liquid cooling solutions. It aims at accelerating Time To Compute and optimizing Tokens Per Watt, which represents AI throughput measured per unit of energy expenditure.
Siemens inherits decades of experience and knowledge about electrical systems and intelligent infrastructure solutions for industrial applications and transfers it into the data center market. Siemens seamlessly integrates IoT-ready hardware, AI technologies, cloud-based software, and comprehensive digital services to enable data center operators to drive change and scale with confidence. Its solution helps deliver reliable, efficient, and sustainable operation for mission-critical facilities with demanding AI workloads.
“This reference solution reduces times-to-compute and delivers optimal tokens per watt performance,” said Ciaran Flanagan, global head of data center solutions at Siemens. “This solution will scale and be fault-tolerant and energy-efficient.”
Industry knowledge leads to innovation in cooling technologies
nVent is a leader and innovator in liquid cooling with an impressive track record of successfully addressing complex cooling challenges across global cloud service providers. Its industry-leading team of experts uses comprehensive product portfolios to enable data center managers to achieve their operational objectives through collaboration with leaders in chip manufacturers and hyperscalers. nVent enables reliable and scalable liquid cooling solutions that address the future demands of high-density computing.
This collaboration embodies a paradigm shift with regard to developing AI infrastructure, with a fusion of engineering expertise and cooling innovation. The collaborative reference architecture enables data center operators to offer streamlined solutions that simplify deployments without losing flexibility. By solving common challenges associated with scaling, cooling, and power density it places the sector on track for healthy growth with regard to AI computing.
