Jelena Pejković, Sales Director of Schneider Electric's Reliable Power Supplies Sector: We Set Standards for AI-ready and Climate-positive Data Centers

- The market of data centers in Europe, as well as in the region, is experiencing strong growth, and current predictions indicate a rate of about 11.9% per year. The main driver of this growth is Artificial Intelligence (AI) – particularly generative models and machine learning. These technologies require massive processing power and specialized GPU clusters, leading to increased density and energy consumption in data centers. The assessment shows that AI can increase energy consumption by up to 20 times per unit of space compared to traditional data centers. In addition to AI, market growth is also fueled by Cloud services, which remain the basis of digital infrastructure, as well as digital transformation in all industries, which generates ever-increasing needs for processing and storing data. Edge computing is gaining importance due to the demand for low latency and local data processing, in particular for IoT and AI applications, which encourages the development of decentralized edge centers. Finally, regulation plays an important role, especially in relation to digital sovereignty. European countries insist on data localization and energy efficiency, which is shaping investment in "sovereign clouds" and setting new standards for sustainability and infrastructure security.

- What we see is an explosion in energy consumption: Generative AI and training large models require enormous computational resources. A typical enterprise rack consumes 7–10 kW, while AI racks exceed 30 kW, and in the case of hyperscale centers, this consumption exceeds 100–300 kW per rack. Instead of CPUs, AI requires GPU clusters, which increases density as well as heat dissipation. New chips (eg NVIDIA Blackwell) have a TDP (Thermal Design Power) over 1000W. Traditional AC systems are no longer enough, nor is traditional air cooling. New densities and new powers require much more, which is why the switch to 800 VDC distribution and liquid cooling is underway.

Artificial intelligence is definitely changing the rules of the game - from power and cooling to architecture itself. Traditional data centers are not designed for these demands, so the industry is moving to modular AI-ready facilities with liquid cooling, DC distribution and advanced energy-efficient solutions.

- Schneider Electric was named the most sustainable corporation in Europe by Corporate Knights, and the most sustainable corporation in the world by TIME magazine and the Statista platform two years in a row. These recognitions confirm that sustainability is built into the very purpose and strategy of the company - from product development to business models. The goal is decarbonization and digitalization as the two key pillars of the energy transition.

- The consumption of electricity in data centers that support AI applications is growing exponentially. According to estimates by the International Energy Agency (IEA), by 2030 it will more than double and reach around 945 TWh, which is approximately equal to the total annual electricity consumption in Japan.

- Schneider Electric and NVIDIA are strategic partners in the development of AI-ready infrastructure for data centers and so-called "AI factories". The aim of the collaboration is to respond to the growing demands for high power density, advanced cooling and sustainability, brought about by the mass application of artificial intelligence.

Schneider and NVIDIA have developed the first publicly available reference designs for AI hubs, optimized for NVIDIA GPU clusters (GB200 NVL72, Blackwell series). Designs include integrated power management and liquid cooling, support for up to 132 kW per rack, and interoperability with NVIDIA Mission Control for AI workload orchestration.

We further discuss the implementation of liquid-to-liquid CDUs and direct-to-chip cooling for extremely dense AI infrastructure. Through the acquisition of Motivair, Schneider Electric has strengthened its liquid cooling portfolio, which reduces energy consumption and increases efficiency.

- Schneider Electric has a key role in the AI ​​ecosystem – from power and cooling to digital control. Our solutions enable GPU data centers and AI supercomputers to be energy efficient, scalable and sustainable, supported by the latest technologies and partnerships with global leaders like NVIDIA.

- The ideal data center of the future is a fully digitized, energy-optimized and sustainable ecosystem, based on modular architecture, liquid cooling technologies and AI-driven management. Key technologies include advanced UPS systems, new generation batteries, edge infrastructure and integration of renewable energy sources. Schneider Electric is already developing such solutions through the EcoStruxure™ platform, modular sub-systems and partnerships with leaders such as NVIDIA, thereby setting the standard for AI-ready and climate-positive data centers.

- AI is a game-changing opportunity, but requires smart and sustainable investments. The focus should be on energy efficiency, modular solutions and the integration of renewable sources. Who invests in AI-ready infrastructure today, will ensure a competitive advantage tomorrow. Schneider Electric brings technology, sustainability and performance together, and what we can offer locally is the expertise of our employees as well as our partners. Also, there is our willingness to get involved with investors and designers, as consultants, at the earliest stage of any project.