Infineon and Skeleton Technologies Partner to Develop High-Efficiency Power Solutions for AI Data Centers

Infineon Technologies AG and Skeleton Technologies, a global provider of high-power energy storage systems widely used in applications such as AI data centers, have signed a Memorandum of Understanding (MoU) to jointly explore next-generation power solutions for modern data center infrastructure.

The rapid expansion of artificial intelligence is driving unprecedented growth in data center power consumption. At the same time, operators are demanding infrastructure that can deliver higher power density, greater efficiency, and exceptional reliability. These trends are accelerating the adoption of new power architectures, with solid-state transformers (SSTs) and high-power sidecar modules emerging as important building blocks for future AI data centers. By improving power conversion efficiency from the grid to the computing core while enhancing system resilience, these technologies can help data centers meet increasingly demanding AI workloads.

Under the new partnership, Infineon and Skeleton Technologies will combine their complementary technologies and expertise to advance high-performance power architectures for AI infrastructure.

“Andreas Weisl, Executive Vice President and Chief Sales Officer of Infineon’s Green Industrial Power division, said: “AI is fundamentally changing the requirements of power infrastructure. As the power density of modern data centers continues to rise, new approaches to energy conversion, distribution, and storage are becoming essential.”

“By combining Infineon’s expertise in power semiconductor technologies with Skeleton’s innovations in supercapacitors and power conversion systems, we are working toward highly efficient and reliable power architectures capable of supporting the next generation of AI workloads.”

Skeleton Technologies CEO Taavi Madiberk highlighted power density as one of the major challenges facing the rapidly expanding AI infrastructure market.
“Power density is becoming one of the key constraints for AI infrastructure worldwide,” said Madiberk. “AI data centers need to deliver more power to increasingly dense GPU systems without allowing the power infrastructure itself to consume additional space.”

He added that the collaboration between Infineon and Skeleton Technologies is designed to address this challenge by combining Infineon’s precision in power conversion and control with Skeleton’s expertise in rapidly storing and delivering extremely high levels of power. This combination could enable higher power density and greater computing capacity within the same physical footprint.

As part of the agreement, the two companies will investigate and develop advanced power architectures covering the entire AI data center power chain, from the electrical grid through to the computing core.

One key area of cooperation will focus on next-generation supercapacitor-based solid-state transformer solutions for converting medium-voltage AC into high-voltage DC. The companies plan to combine Infineon’s high-efficiency, high-voltage CoolSiC™ power semiconductor technology with Skeleton’s capabilities in power conversion systems and supercapacitor technology.

In addition, Infineon and Skeleton are collaborating on a GaN-based peak-shaving system that integrates Skeleton’s advanced supercapacitors with Infineon’s CoolGaN™ power semiconductors. Such an architecture can provide rapid power delivery during peak demand while helping reduce stress on the wider power infrastructure.

By bringing together complementary semiconductor, energy storage, and system-level technologies, the partnership aims to improve the energy efficiency, power density, and reliability of next-generation AI data centers. At the same time, the solutions could help operators optimize infrastructure utilization and reduce the total cost of ownership as AI computing requirements continue to grow.