According to Bloomberg, the surge in artificial intelligence (AI) is providing a significant boost to some of Europe's largest industrial companies, helping them navigate weaker manufacturing demand from regions such as Asia and North America. Data center revenues for the six biggest electrical firms reached €20 billion ($22.2 billion) last year, doubling from five years earlier, as reported by Redburn Atlantic analyst. The AI boom is unprecedented in the data center industry over the past two decades, with AI chips consuming more power and generating more heat than other applications, necessitating advanced cooling technologies.
Companies like Schneider Electric and Siemens are at the forefront, supplying the necessary equipment and technology to prevent overheating in high-density servers. The increasing footprint of AI workloads is expected to drive a multi-year wave of investments in data center and server-cooling technologies, benefiting companies like Vertiv and Schneider. Schneider Electric, which derives 17% of its revenue from data centers, emphasizes cooling technologies, particularly liquid cooling, which is more effective than air cooling in high-density environments. Schneider Electric's Sharma highlighted the unprecedented deployment needs compared to previous cycles such as e-commerce, Bitcoin mining, or the metaverse.
Siemens is also seeing benefits from orders related to its low- and medium-voltage electrification infrastructure and cooling systems, including temperature sensors and flow meters. Ciaran Flanagan, Siemens' global head of data center solutions, noted that a significant portion of their current opportunities is AI-related. Despite recent investor concerns about the timeline for returns on AI investments, Siemens' Chief Executive Officer remains optimistic, citing the growing amount of data and computational needs. This growth is expected to drive the electrification market into a super cycle, which is crucial for Siemens, especially given the recent weakness in its automation business.
However, challenges remain, as demand for power continues to outstrip supply, and planning permission delays are hindering data center development. These bottlenecks need to be addressed to fully realize the earnings potential from building AI infrastructure. The market thesis for companies involved in AI infrastructure is well understood, but the duration of this growth phase remains a topic of debate, as noted by a Citi analyst.