Ambarella (AMBA) 28th Annual Needham Growth Conference Virtual summary

Company transformation and AI product evolution

• Transitioned from video processors for consumer applications to AI SoCs for edge and infrastructure markets over the past decade.

• Developed three generations of proprietary AI accelerators, with the latest focusing on transformer-based models and supporting advanced GenAI applications.

• Edge AI now accounts for approximately 80% of annual revenue, with over 40 million AI SoCs shipped and more than $1.3 billion invested in AI R&D.

• ASPs for new chips range from $20 to $400, up from $15 to $75 for previous generations, with edge infrastructure products like N1-665 featuring three-digit ASPs.

• AI is now embedded in all new products, with a differentiated "AI Algorithm First" approach for efficiency and scalability.

Product and technology highlights

• Announced CV7 chip with 2-2.5x AI performance and lower power than previous generation, and introduced the first 4nm and 2nm AI SoCs.

• Third-generation chips support models from 500 million to 100 billion parameters, with low-end chips running 2 billion-parameter models at 2W.

• Demonstrated 31 new edge AI applications at CES 2026, showcasing the open Cooper Developer Platform and expanding Developer Zone.

• Continued investment in advanced nodes (down to 2nm) to improve performance per watt and enable new applications.

• Platform enables high programmability and flexibility, supporting a wide range of customer applications and markets.

Market opportunities and applications

• Transformer-based chips open new markets in robotics, autonomous vehicles, drones, and edge infrastructure.

• GenAI applications demonstrated in retail and hospitality, enabling advanced analytics and marketing data collection from existing security cameras.

• IoT business diversifying, with security now less than 50% of segment revenue; growth in drones, wearables, and portable video.

• Edge infrastructure products aggregate multiple endpoints, upgrading legacy devices with AI capabilities.

• AI is expanding use cases in enterprise security, smart home, automotive, robotics, and edge infrastructure.