Schneider Electric has partnered with Nvidia to launch a reference design aimed at facilitating the deployment of generative AI within traditional data centers.
Generative AI is experiencing rapid growth, often referred to as a “digital renaissance” comparable to the Internet boom of the 1990s. According to Stanford University’s AI Index report, the impact of AI is expected to increase significantly, with investment in 2023 reaching $25.2 billion—nine times higher than in 2022 and 30 times more than in 2019. This surge in AI development presents a significant opportunity for data centers to expand their services to meet growing AI demands. However, such expansion and innovation come with high costs, and there is an ongoing need to enhance energy efficiency and ensure sustainable operations.
Generative AI models rely on specialized chips, with Nvidia leading the market in GPU technology. Traditional data centers, however, are not typically equipped with the power, cooling, and racking systems necessary to support the high-density deployment of these advanced GPUs. To address this challenge, Schneider Electric has collaborated with Nvidia to create reference designs for AI data center infrastructure.
These reference designs serve as a foundational blueprint, outlining performance specifications, materials, and components, which can be tailored to meet the specific needs of each data center.