onsemi bets $600M on Ferric to power AI chips
What's the deal? onsemi has agreed to acquire Ferric, a developer of integrated voltage regulator (IVR) technology, for $600 million in cash plus up to $175 million tied to revenue milestones. The chipmaker says the deal will strengthen its position in power delivery for AI processors, with the acquisition expected to close by the end of 2026.
What does Ferric do? The company integrates the power controller, FETs, inductors, and capacitors into a single compact package. onsemi says the architecture cuts board area by 50 to 70% versus conventional discrete regulators while delivering more power closer to the processor. Ferric brings more than a decade of engineering expertise and over 40 patents.
Why now? As processor power scales, the industry is moving voltage regulation progressively closer to the chip to reduce losses and respond faster to shifting demand. IVRs integrate power conversion directly into the processor package or substrate — the next step in that shift.
What's the endgame? onsemi raised its 2030 AI data center revenue target by $500 million to more than $3 billion after announcing the deal. Its AI data center revenue reached $180 million in Q3 2026.
"The rapid growth of AI is reshaping the requirements for power delivery to increasingly capable and power-intensive processors," said Hassane El-Khoury, onsemi's president and chief executive officer.
Ferric founder and chief executive officer Noah Sturcken said joining onsemi gives the startup "the scale, manufacturing capabilities and customer reach needed to accelerate adoption."
What could go wrong? onsemi says Ferric's near-term revenue contribution will be limited. The payoff hinges on whether Ferric's IVRs win design slots as processor power demands rise and deliver the promised efficiency and space savings at scale.
The signal: The deal fits a clear push to bring power conversion closer to AI processors. In August, Navitas agreed to buy Claros, a vertical-power-delivery developer, for up to about $233 million — a sign of growing competition in this part of the AI infrastructure stack. The $600 million upfront price marks a sizable strategic bet on a technology whose significance is still to be proven in the market.
Read more: globenewswire.com
Image credit: FritzchensFritz
