SAP CEO rebukes Nvidia’s AI infrastructure push

SAP Chief Executive Officer Christian Klein has bristled at the notion that Europe needs to catch up with the rest of the world by constructing more data centers filled with high-powered chips. 

Speaking to reporters at a press call from SAP’s headquarters in Walldorf, Germany, Klein singled out comments made by Nvidia CEO Jensen Huang during his headline-grabbing European tour.

“Is it really that we need to build five data centers and put great chips in there? Is this what Europe needs? I doubt it,” Klein said.

Huang, whose company is the market leader for AI chips, declared at an event in June that Europe is losing ground in artificial intelligence because it doesn’t have enough computing power. 

During his visit, Huang announced several partnerships to help build that AI infrastructure on the continent, all predicated on Nvidia’s technology. He called on European leaders to take the kind of plunge that US tech giants were doing, investing huge amounts across many sectors in more data processing power.

However, Klein argues that this approach would be a mistake for Europe. For him, the continent’s power is its industrial base, automotive, manufacturing, chemicals, engineering, and so on, not the ability to go head-to-head with American tech titans in the race to the bottom of raw computing.

Instead of trying to emulate Silicon Valley’s hardware-first strategy, Klein said Europe should focus on creating clever AI applications that serve its core industries. 

SAP steps back from gigafactory ambitions

Klein’s statements indicate a drastic change in SAP’s AI strategy. Just six months ago, he said he fully supported Europe building its own AI “gigafactories” — massive facilities for training and scaling AI models. At the World Economic Forum in Davos in January, he cited the US-led “Stargate” effort as a “great role model” for Europe. He vowed to “absolutely support” a similar European effort at the time.

SAP, behind the scenes, had been in discussions with other German companies to make a joint investment in one of a series of European Union-backed AI gigafactories. But the discussions eventually broke down. In the words of SAP’s head of customer services and delivery, Thomas Saueressig, it turned “into nothing” because of divergent visions and a short-term view.

This week, SAP clarified that it is no longer interested in investing in or participating operationally in such mega-projects. Instead, it wants to back such initiatives as a tech and software provider, supplying tools, platforms, and customisable AI applications for those doing infrastructure.

Earlier this year, the European Union committed to spending 20 billion euros ($23 billion) to create five AI gigafactories. If the ambition is clear, SAP’s retreat to the sidelines underscores an increasing skepticism within Europe’s industrial leaders about trying to copy the American model of AI supremacy.

Developers embrace open-source AI for faster innovation

Klein also wondered about the value of racing in the long term to train huge proprietary models of AI. Instead, he sees a move toward cheaper, open-source alternatives. He cited as an example the Chinese company DeepSeek, which recently made headlines for besting the top US AI companies by using a model constructed on open-source code trained at a fraction of the cost.

Klein said the success of open-source models demonstrates that it isn’t necessary to have unlimited computing power or billion-dollar budgets to develop effective AI solutions. He emphasized that what matters are intelligent, streamlined models that are relevant to the market and function well in real business environments.

He proposed that the days of monolithic AI infrastructure are waning. Now that models are mainstream, the significant issue will be how effectively companies apply these tools to real-world problems.

SAP’s priority is to integrate AI capabilities into enterprise software systems. Klein has written about AI-augmented workflows for logistics, predictive maintenance for factory equipment, intelligent customer service bots, and real-time decision-making dashboards for supply chain managers.

Cryptopolitan Academy: Coming Soon - A New Way to Earn Passive Income with DeFi in 2025. Learn More