How Schneider Electric Leads the AI Data Centre Revolution
The data centre industry stands at a crossroads. As AI workloads demand unprecedented levels of computing power, data centre operators must engage more with sustainability imperatives to achieve more efficient operations.
This has led to a tension between innovation and environmental responsibility, which has quickly become the defining challenge of 2025: how can the data centre industry meet soaring AI demands without compromising sustainability commitments?
For Schneider Electric, this challenge represents an opportunity to demonstrate its extensive leadership in sustainable technology solutions. The French multinational corporation, which specialises in energy management and automation, is globally recognised for being at the forefront of this transformation: so much so that it was named the most sustainable company by TIME in June 2025.
Nirupa Chander, Senior Vice President for Secure Power & Data Centres International at Schneider Electric, oversees the company’s data centre business across emerging markets. Based in Dubai, she leads regions within the International team with the exception of Europe, the US and China.
The scope of her responsibility reflects the global nature of the AI revolution, as events from the rise of DeepSeek to US tariffs have impacted the wider industry.
“We still have many markets that haven’t got their own cloud and AI infrastructure in place,” she explains. “We see a continued overall positive outlook. It’s been a tumultuous six months in the industry, but what fundamentally hasn’t changed is the overall outlook in the demand for digital infrastructure across the regions.”
How power demand is reshaping infrastructure requirements
AI continues to change rapidly as adoption rises.
For instance, ChatGPT acquired its first 100 million users in just two months, compared to nine months for TikTok. Likewise, the market for Gen AI is expected to hit US$1.3tn by 2032, according to Bloomberg, which PwC says could add up to US$15.7tn to the global economy in 2030.
So in just the last few years alone, surging demand for AI has to be served by the physical infrastructure – data centres – which take time to build.
“That’s where we come in because we’re a deployer of physical infrastructure to enable a lot of this AI demand that we see in user usage and applications,” Nirupa explains. “The market is overall very dynamic and still positive.”
Despite a very interesting start to this year, Gen AI is hitting US$1.3tn by 2032 in terms of market value. However, Nirupa explains that the other challenge with AI concerns energy consumption.
Nirupa adds that where Schneider Electric’s expertise comes in is in relation to energy demand: “Energy density for AI infrastructure is very high – up to 10 times for every single query which needs to be served. Power availability and capacity and how that energy is most efficiently deployed when you’re building this vertical infrastructure is a key topic as well.”
As it looks ahead, Schneider Electric is looking to balance sustainability with growing power demands, especially when it comes to data centres and AI. This is particularly important, as average rack densities have moved from 10-15kW to more than 100kW per rack – an extreme transition in a short space of time.
“It requires a lot of technology to help deliver the power and cooling requirements for these machines to work,” Nirupa says. “That’s where Schneider comes in. We’re looking at how energy consumption in a data centre can be reduced. So what we’re telling the market is that it’s not only energy for AI, but it’s also how you can use AI to improve energy efficiency within data centres.
“We’re experts in power and cooling and are providing technology solutions from Chip to Chiller.”
Accommodating both traditional and emerging workloads
Transitioning to a new approach within the data centre will require hybrid architectures, particularly when it comes to inference applications. With this in mind, Nirupa explains that the data centre industry is going to see a rising demand for more edge applications and Schneider Electric suspects most will be hybrid architectures.
“Across multiple markets, we’re seeing more applications at the edge, so closer to where the demand is,” Nirupa says. “So, while data gets trained in these mega data centres in the US or in other locations, the inference applications tend to be typically closer to where the applications are being used.
“To confront this, we are delivering solutions for both air and liquid hybrid cooling systems to help manage the thermal load of these data centres.”
She adds: “It’s about managing the thermal load generated by AI servers and chips, but also about how you save every single watt of power or cooling that is possible to ensure the most efficient way to build AI infrastructures - that are then being used for delivering productivity gains, improving quality of life, doing things like cancer research and supporting some very mission-critical applications.”
Schneider Electric has doubled down on its commitment to sustainability and now wins multiple awards a year, including the recognition from TIME, for its environmental strategy. A significant part of this is its liquid cooling solutions, which help data centres meet their sustainability goals.
“Liquid cooling is the most efficient way to cool this level of thermal load that we're seeing in AI compute now,” Nirupa explains. “High workloads generate a lot of heat and need a lot of power. The acquisition of Motivair marks an important step as its unique liquid cooling portfolio completes our Chip to Chiller portfolio bringing decades of experience in AI together with Schneider Electric’s sustainability expertise for our customers.”
The Motivair acquisition closed at the start of 2025, with Schneider Electric now harnessing its liquid cooling solutions as part of its portfolio to support continued AI growth.
“Motivair completes our solution portfolio around AI and Chip to Chiller for data centres.”
Cooling distribution units (CDUs), are also working to manage heat load transfer from the IT load and enhance its liquid cooling solutions for customers. In addition, Nirupa explains that Schneider Electric has modular solutions and reference designs that prioritise efficiency. There are a number of white papers to help customers design the topology that’s right for them more sustainably.
“With these solutions, what's important is to understand the overall design of the system,” she says.
Schneider Electric is partnering with chip manufacturers like NVIDIA, having announced at GTC Paris that they would be working together to power AI Factories. Both companies are eager to leverage their shared experience in AI-ready infrastructure, sustainability and grid coordination to respond to the European Commission’s ‘AI Continent Action Plan’ – an initiative that outlines a shared mission to set up at least 13 AI factories across Europe, while establishing up to five AI gigafactories.
“We're working on reference architectures in terms of how to deploy high-powered machines with high cooling requirements in a standardised design architecture at scale,” Nirupa explains. “Our partnership with NVIDIA is very important because it's a technology partnership. As technology changes very fast, it requires the industry to work together to build solutions that can be deployed at scale and at speed.
“We're also working with internet giants, cloud providers, universities and academia partnerships to make sure that we're educating the industry and the up-and-coming generation in terms of how things are changing.”
Staying competitive as a digital transformation leader
Looking ahead, as data centres face greater pressures on account of AI, Schneider Electric remains conscious of its position to bridge the gap between innovation and sustainability.
“The key is always staying a step ahead in terms of how technology is evolving in the IT space,” Nirupa says. “This is critical for us to deliver solutions to meet future demand.”
She credits the company’s partnerships like those with NVIDIA and other manufacturers as essential in charting how AI architectures are changing.
“You can see that a lot of the new architectures are moving from alternating current (AC) to direct current (DC) because it’s more efficient, which means that the way the electrical infrastructure needs to be designed is changing as well,” she adds. “This means we have to keep up with what’s happening with server manufacturers and the chip manufacturers to stay ahead of the game.”
As AI moves towards inference workloads, data centre design too is expected to evolve.
“We can see architectures will become more decentralised, closer to where the demand is,” Nirupa adds. “We therefore expect edge data centres to become more popular as demand for overall infrastructure continues to grow.
“Likewise, modularity in design so that infrastructure can be deployed at speed and at scale is another key change – and that is how we’re evolving our portfolio as well.
“We are also making sure that we’re delivering products that drive the highest level of efficiency, whether it’s on the cooling side or on the power side. One example is that Schneider Electric has just released the world’s most efficient UPS (uninterruptible power supply), which has the highest rate of efficiency and the lowest footprint in the market. Our new Galaxy VXL UPS combines a compact, innovative, and highly efficient design with enhanced safety features, providing world-leading power protection for a wide range of AI, data centre and industrial applications. This ensures the high density workloads of the future are supported by unparalleled levels of reliability.”
She adds: “It’s about having efficiency across design, architecture and at product level as AI workloads continue to demand more energy.”
