Only 22% of Ambarella's revenue is cars, and the rest is cameras, robotics, and AI that runs on the device, which is reportedly what NXP is paying for.
NXP Semiconductors is reportedly in talks to acquire Ambarella, a Santa Clara chip designer most consumers have never heard of unless they own a dashcam. The Financial Times broke the story and industry coverage framed it as auto-chip consolidation. Ambarella is sold as an automotive chipmaker, but only about 22% of its revenue is cars. The other ~78% is security cameras, robotics, and on-device AI inference, and that is the asset NXP is reportedly after.
NXP is a microcontroller and connectivity heavyweight with a deeply automotive book of business. Roughly 58% of its semiconductor revenue is automotive, and another 19% sits in industrial and IoT, per an EE Times analysis of the company. As cars turn into software-defined vehicles, the value is moving from the dashboard controller toward the sensors themselves. Cabin software is updated over the air rather than frozen at sale, and cameras and radar need to run AI models locally, on a few watts, before any central computer sees the data.
Ambarella is built for that job. Founded in 2004 and public since 2012, the company has spent two decades on image-processing silicon, starting with classical computer-vision designs in ADAS (advanced driver assistance systems, the camera-and-radar stack behind automatic emergency braking and lane keeping) and security cameras, then pivoting to AI-based computer vision. Specific workloads like driver-monitoring cameras, 360-degree parking views, and traffic-sign recognition all run on the camera's own silicon before any central computer sees the frame, and that is the layer Ambarella's CV chips target. Its current products are low-power, camera-centric system-on-chip (SoC) designs that run convolutional neural networks on-device instead of uploading frames to a cloud server. NXP has the microcontroller footprint in the car but does not have a serious on-device AI camera line of its own. Buying Ambarella plugs that gap and extends NXP into non-auto edge-AI markets, the security and robotics side where Ambarella is currently the stronger silicon house.
EE Times cites the mid-2025 onsemi acquisition of Synaptics as the analog: an incumbent semiconductor company picking up edge-AI silicon it did not build organically. NXP's reported approach to Ambarella fits the same template, applied to the camera-and-AI layer. SiliconAngle and Global Banking & Finance both confirm the FT-sourced talks, though neither adds new mechanism beyond the original report.
The deal would only read as auto consolidation with a small edge-AI adjacency if NXP folds Ambarella's ADAS line into its existing automotive microcontroller franchise and treats the security-camera business as a divestable side asset. The mix numbers in EE Times' analysis point the other way. Ambarella's automotive share is small, its camera-AI share is large, and NXP already ships the microcontroller tier that an Ambarella SoC would sit next to in a software-defined vehicle. Buying into cameras and on-device AI is the bigger lever, even if the car is the visible beachhead market.
Three things remain unproven. Deal terms and price have not surfaced. The product-fit question, whether Ambarella's SoC roadmap dovetails with NXP's automotive customer pipeline, is open until an integration plan is public. The regulatory path through Beijing, Brussels, and Washington for a cross-border automotive chip deal in 2026 is not a rubber stamp. The deal is reported, not done, and whether NXP is buying edge AI for the next car generation or adding another automotive logo to its roster will only resolve when a term sheet does.