Semicon Briefing

Infineon acquires C2i Semiconductors to expand its innovation capabilities in AI datacenter power management – another strategic acquisition following the completed ams-OSRAM sensor division takeover. The dual strategy of share buybacks and targeted acquisitions demonstrates Infineon's ambitions as a European semiconductor champion in the AI infrastructure segment.

Intel has entered into a pact with Tesla and SpaceX for Musk's planned $25 billion Terafab facility, while ASML's EUV order book through 2027 is fully allocated to TSMC, Samsung, and Intel. In the short term, Terafab remains a long-term project without immediate production readiness, but the alliance signals a new era of industrially defense-oriented chip production in the US.

SEMICON Taiwan 2026 focused on advanced packaging: The actual AI chip bottleneck no longer lies in transistor density but in the interconnect wires between chips – a paradigm shift for the entire industry. TSMC, Merck, and materials specialists presented new approaches to system integration, further increasing investment pressure on packaging equipment.

Samsung Electronics and SK Hynix are evaluating manufacturing equipment from Chinese manufacturer AMEC as a hedge against US export control risks for their Chinese facilities. The move illustrates how Western supply chain policies drive Korean chipmakers into a strategic dilemma between US dependency and Chinese localization.

Following the Tata fab, India is planning a second chipfab with a minimum $2 billion investment by 2031, focusing on quality and supply security rather than mere volume. India is thus positioning itself as another pole in the global chip geography reorganization alongside the US, Europe, Japan, and Taiwan.

Navitas Semiconductor acquires Claros for up to $232.8 million, expanding its portfolio with vertical power delivery technology (VPD) and integrated voltage regulators specifically for AI data centers. The deal underscores growing M&A pressure in the power semiconductor segment as data center operators seek more energy-efficient solutions for GPU clusters.