Powering artificial intelligence requires enormous amounts of computing power and energy, making every percentage point of efficiency matter. For Infineon Technologies, one of the world’s leading semiconductor companies in power systems and the Internet of Things (IoT), the company’s challenge is not just supplying the innovative technologies that enable AI. It is using AI throughout its own operations to become more efficient and more innovative.

“AI is already everywhere — in the office, in the factory, in the smartphone, in the car,” says Alexander Gorski, Chief Operating Officer of Infineon Technologies. “The more effectively we apply AI in our own operations, the better positioned we are to scale the technologies our customers need for their AI applications.”

For Gorski, these priorities reinforce one another. As AI becomes more deeply embedded across industries, customers require semiconductor technologies capable of delivering higher performance while consuming less energy, particularly as tailor-made generative, agentic, and physical AI advance.

Power semiconductors therefore play an increasingly important role in AI infrastructure — from hyperscale data centers to industrial applications. “Our power semiconductors and analog/mixed-signal components help convert, control, and distribute energy from the grid to the processor core with greater efficiency, reliability, and intelligence,” he says.

Making that possible requires continued advances across semiconductor technologies. Infineon’s semiconductor portfolio increasingly draws on silicon carbide (SiC) and gallium nitride (GaN). Gorski notes that Infineon is the first company in the world to manufacture GaN on 300-millimeter wafers at scale — a breakthrough that significantly enhances production and reduces manufacturing costs. These materials reduce energy losses, increase power density, and enable smaller, more effective power systems. Gorski also points to advanced packaging technologies and deep system-level expertise as important components in improving the performance and reliability of AI infrastructure.

Infineon follows what Gorski describes as a hybrid manufacturing model, where investments in in-house production that provide a clear competitive advantage are complemented by external manufacturing partnerships in areas where product differentiation comes primarily from design and software. Gorski says Infineon has the largest production capacity for power semiconductors in the market, allowing production to scale alongside customer growth.

Recent investments illustrate that strategy. The company’s newly opened Smart Power Fab in Dresden significantly expands Infineon’s 300-millimeter manufacturing capacity for semiconductors supporting automotive, industrial, and AI-related applications. Closely linked with its manufacturing site in Villach through its “One Virtual Fab” approach, the facilities enable faster process transfers, faster production ramp-up, and greater flexibility in responding to changing customer demand.

Within manufacturing, using AI to optimize wafer flow through production facilities helps to prevent bottlenecks in much the same way that intelligent traffic management systems keep vehicles moving, says Gorski. AI also supports quality control, predictive maintenance, process optimization, energy management, and knowledge management across the company’s operations.

“We are currently piloting the use of humanoid robots, working to further scale the use of digital twins in development and manufacturing, and increasingly deploying AI agents throughout our value chain,” he says.

The company is extending that digitalization into its manufacturing facilities as well. At its Smart Power Fab in Dresden, Infineon used a digital twin to optimize the building and machine layout before construction, while AI algorithms support system and process qualification as production ramps up.

For Infineon, these internal applications are not separate from its customer-facing technologies. They directly support the ability to innovate and scale manufacturing alongside the rapidly expanding AI market.

“I have never experienced growth dynamics like those now emerging with the AI boom,” he says.

“Existing power conversion architectures in data centers are hitting physical limits — we need entirely new solutions,” says Gorski. “It is exciting to be part of this development and actively shape the way forward.” For Infineon, the next stages represent both a long-term strategic focus and great technical opportunities.