Ambarella Drives AI to the Edge, Expanding Beyond Camera Chips
US-based Ambarella, historically a leading supplier of camera chips, is strategically shifting its focus towards processing artificial intelligence directly on devices – at the “Edge” – rather than relying on cloud-based solutions. This move reflects a broader industry trend towards localized, efficient AI processing.
Evolution from Broadcast to Edge AI
For over two decades, Ambarella has evolved from developing broadcast encoders and low-power video chips to creating AI-enabled sensors for a diverse range of applications, including security, robotics, and industrial automation. Jerome Gigot, VP of Marketing, Edge AI at Ambarella, outlined this strategy at embedded world 2026 [IoT Insider].
CVflow: Algorithm-First Processing
Central to Ambarella’s approach is its proprietary CVflow processor, which integrates computer vision and perception software for both enterprise and consumer devices. The company has shipped over 400 million system-on-chips (SoCs), with more than 42 million being AI-enabled SoCs used in cameras, robots, and video conferencing systems [IoT Insider]. Gigot emphasizes that Ambarella differentiates itself from GPU manufacturers by prioritizing algorithms optimized for real-world perception tasks over general-purpose computing [IoT Insider].
“Physical AI” and Expanding Markets
Ambarella defines “physical AI” as artificial intelligence that interacts with the physical world through sensors like cameras, LiDAR, radar, and microphones [IoT Insider]. While security cameras remain its largest market, the company is actively expanding into sectors like retail, logistics, and manufacturing, where AI applications include shoplifting detection, queue monitoring, and defect detection on assembly lines.
The Benefits of Edge Computing
Edge computing is a core tenet of Ambarella’s strategy. Processing data locally, either on the device itself or on a nearby Edge box, offers several advantages over cloud-based solutions: reduced latency, lower costs, and enhanced security [IoT Insider]. Gigot states that “Edge AI is more efficient, secure, and cost-effective than Cloud processing” [IoT Insider]. Ambarella employs a hybrid model, utilizing on-device AI for immediate tasks, Edge boxes for medium workloads, and cloud computing for large-scale analysis.
Financial Growth and Developer Ecosystem
Edge AI currently accounts for 80% of Ambarella’s revenue, representing a 50% growth in this segment over the past year [IoT Insider]. The company is also fostering a developer ecosystem through its newly launched Developer Zone, providing software development kits (SDKs), pre-trained models, and low-code/no-code tools to enable third-party developers to create specialized solutions.
Future Developments
Ambarella is developing a next-generation chip based on a 3-nanometer process, promising increased AI performance with reduced power consumption, crucial for Edge devices. The company is also continuing to expand its camera product line and software ecosystem.
Ethical Considerations
Ambarella acknowledges the ethical implications of its technology and incorporates features like face-blurring masks into its chips. The company also restricts sales to certain countries, but recognizes its limited control over how its products are ultimately used.
Jerome Gigot summarizes Ambarella’s position: “We’ve been doing physical AI for 20 years,” he said. “Others are moving from Cloud to Edge. We started from the real world and are building upwards.” [IoT Insider]
Jerome Gigot is the Vice President of Marketing at Ambarella [Ambarella] and has over 20 years of experience in video technology [LinkedIn].