Tech news in 3 minutes
Gartner names Infineon as company to beat in AI data-center power semiconductors
Gartner has named Infineon Technologies AG of Munich, Germany as "the company to beat" in the rapidly evolving market for AI data-center power semiconductors, according to a report issued on 18 May by analysts Menglin Cao and Rajeev Rajput. The report cites Infineon's comprehensive product portfolio, manufacturing capabilities, and early investment in advanced technologies as key factors behind its front-runner status. However, Gartner notes that growing competition in silicon carbide (SiC) and gallium gallium nitride (GaN), as well as rivals focused on the compute board level, are challenging Infineon's position—challenges the company is addressing through its broad portfolio and system-level expertise. Infineon projects €2.5 billion in AI market revenue for its fiscal year 2027, underpinned by what it says is the leading portfolio position across the AI power delivery chain. Its "grid-to-core" solution spans the entire power architecture, from solid-state transformers and power supply units to processor-level power management, addressing the unprecedented power density and thermal bottlenecks of high-performance AI workloads. By optimizing every conversion stage with the right semiconductor technology—SiC for high-efficiency, high-voltage grid-to-rack conversions, GaN for ultra-dense, high-frequency intermediate stages, and silicon at the processor level—Infineon minimizes cumulative energy losses, a critical advantage at hyperscale. Beyond data centers, Infineon is extending its solutions to physical AI, including humanoid robots and autonomous systems. With the global robotics market projected to reach up to $1.7 trillion by 2050 and 300 million humanoids expected by then, and an average semiconductor bill of materials of $500 per humanoid, physical AI represents a major growth opportunity. Infineon’s expertise in microcontrollers, power systems, sensing, connectivity, functional safety, and security positions it to power the full arc of the AI era.
View original article
2026-07-21
See Defacto Technologies at DAC 2026