Navitas to buy Claros for up to $232.8M in AI power-chip push
What's the deal? Navitas (Nasdaq: NVTS) has signed a definitive agreement to acquire power-management firm Claros in a deal valued at up to approximately $232.8 million, based on Navitas' August 21, 2026 share price. Claros develops vertical power delivery (VPD) and integrated voltage regulator (IVR) technology for AI data centers.
Why now? AI data centers demand ever-more efficient power conversion, from the grid down to the processors. Navitas says the acquisition completes its "grid-to-xPU" high-power portfolio, converting ultra-high-voltage grid power all the way down to AI chips.
What's the endgame? Navitas expects the deal to more than double its 2030 serviceable addressable market to over $8 billion, adding at least $3.5 billion from VPD and IVR technology. It also projects revenue acceleration and margin expansion from 2028 to 2029 onward, without changing its current path to profitability under the Navitas 2.0 strategy.
The terms: Roughly $216.0 million would be paid at closing in cash and Navitas stock. The remainder — up to $16.8 million — is deferred, paid in shares only if specific business milestones are met in the two years after closing. Separately, certain continuing Claros employees are eligible for about $28.9 million in performance-based Navitas shares tied to the same milestones.
What could go wrong? The deal is signed but not closed, pending regulatory and customary conditions, with a target of before year-end 2026. Issuing consideration shares would increase Navitas' total share count and dilute existing holders. The company has not disclosed the cash-versus-share split or number of shares, so the precise ownership effect is unclear.
The signal: The purchase reflects the race among power-chip makers to own the full electrical path into AI compute, where energy efficiency increasingly determines data-center economics. By stitching together grid-to-processor conversion in one portfolio, Navitas is betting that power delivery — not just the chips themselves — becomes a central battleground in AI infrastructure.
Read more: stocktitan.net
Image credit: Generated with Gemini
