Ambarella is a leading developer of low-power system-on-a-chip semiconductors and software for edge and physical artificial intelligence applications and intelligent automation. The company designs products that enable electronic systems to become partially or fully autonomous through features such as intelligent automation complex scene understanding and autonomous decision-making. Ambarella's technologies perform multi-modal data fusion and complex data analysis in real…
Ambarella is a leading developer of low-power system-on-a-chip semiconductors and software for edge and physical artificial intelligence applications and intelligent automation. The company designs products that enable electronic systems to become partially or fully autonomous through features such as intelligent automation complex scene understanding and autonomous decision-making. Ambarella's technologies perform multi-modal data fusion and complex data analysis in real time delivering high quality imagery while preserving vital system resources like power and network bandwidth. The company specializes in deployable scalable designs for intelligent electronic systems that utilize high-bandwidth sensors offering a proven path to mass production. Its products are used in a wide variety of applications including video security cameras automotive camera systems fixed robots autonomous mobile robots industrial applications intelligent transportation systems and consumer devices such as action drone and 360° cameras.
Ambarella generates revenue primarily through the sale of its system-on-chip products and related software solutions. The company's revenue comes from its AI inference SoCs that integrate video processor technology with its proprietary CVflow deep learning neural network processor. These products are sold to original design manufacturers and original equipment manufacturers globally. Ambarella's customer base is concentrated in Asia with sales to customers in that region accounting for approximately 88% 85% and 79% of total revenue in the fiscal years ended January 31 2026 2025 and 2024 respectively. The company relies on a limited number of customers for a significant portion of its revenue with WT Microelectronics Co Ltd formerly Wintech Microelectronics Co Ltd accounting for approximately 70% of total revenue in fiscal year 2026.
The company operates through the following segments:
• No operational segments are reported in the filing. The company does not disclose separate operating segments for financial reporting purposes.
Ambarella holds a leading position in the edge AI inference market particularly for physical AI applications that require real-time perception and reaction to the environment. The company competes against large international semiconductor companies such as HiSilicon Technologies Co Ltd Novatek Microelectronics Corp NVIDIA Corporation Qualcomm Incorporated and Sigmastar Technology Ltd in the IoT market. In the automotive camera market Ambarella faces competition from Horizon Robotics Inc Mobileye a subsidiary of Intel Corporation Novatek NVIDIA Qualcomm Renesas Electronics Corporation and Texas Instruments. Ambarella's competitive strengths include its proprietary CVflow AI processing architecture high-performance low-power AI video and image algorithm expertise deep sensor fusion capabilities highly integrated SoC solutions based on a scalable platform comprehensive and flexible software advanced VLSI capabilities and a strong developer ecosystem. These strengths enable the company to deliver AI performance power efficiency and functional integration levels required for next-generation edge and physical AI applications.
Ambarella serves a diverse customer base including original design manufacturers and original equipment manufacturers across multiple industries. The company's products are used in enterprise and home security camera markets portable video applications automotive applications including advanced driver assistance systems cabin monitoring systems AI telematics systems electronic mirrors and automotive video recorders autonomy applications such as central domain controllers for L2+ to L4 autonomous vehicles and emerging robotic applications including fixed and mobile robotics. Other IoT applications include video conferencing access control healthcare medical and virtual reality applications. Ambarella's technology enables edge AI boxes that aggregate multiple video audio and other modality streams of data and provide local generative AI inferencing through vision-language models and large language models.
Sector:TechnologySector rationaleAmbarella designs and sells system-on-a-chip (SoC) semiconductors and related software for AI inference and intelligent automation. Its revenue is derived from selling these high-tech components to original equipment manufacturers (OEMs) and original design manufacturers (ODMs), which fits the Semiconductor Design and AI Platforms industries within the Technology sector.Industries:Semiconductor DesignTechnologyPrimaryAmbarella designs low-power system-on-a-chip (SoC) semiconductors, including AI inference SoCs and the proprietary CVflow deep learning neural network processor. The company operates as a fabless designer, selling these chips to original design manufacturers (ODMs) and original equipment manufacturers (OEMs).AI PlatformsTechnologySecondaryThe company sells AI inference capabilities as a standalone product through its AI inference SoCs and software solutions that enable local generative AI inferencing using vision-language and large language models.Autonomous DrivingTechnologySecondaryAmbarella provides the software and hardware stack for autonomous vehicles, specifically central domain controllers for L2+ to L4 autonomous vehicles and advanced driver assistance systems (ADAS).Classified using BQ-MICSCIK: 0001280263
Investment Thesis
▲ Bull case
Ambarella's strategic positioning as one of the few companies capable of integrating all edge accelerated computing functions—CPU, AI accelerator, ISP, encoding, and system control—into a single SoC provides a durable competitive moat as AI workloads evolve toward GenAI and Agent AI at the edge. The company's CVflow architecture supports over 200 AI model architectures in production, and its unified software SDK allows seamless portability across its product stack, reducing customer integration costs and enabling faster time-to-market for complex applications like vision-language-action models in robotics and multimodal reasoning in industrial automation. This full-stack expertise is increasingly rare and valuable as customers seek to avoid fragmented multi-vendor solutions in data centers, creating a structural advantage that is difficult to replicate and positions Ambarella as a critical enabler in the physical AI and embodied AI ecosystems where real-time, low-latency processing is essential.
The execution of long-term agreements (LTAs), exemplified by the Hanwha deal with potential revenue exceeding $800 million over 10+ years, represents a transformative shift toward predictable, multigenerational revenue streams that extend beyond traditional transactional sales. These LTAs involve co-development of semi-custom SoCs, NRE funding from customers, and joint technology roadmap planning, which not only secure wafer supply commitments through partnerships with Samsung but also expand Ambarella’s addressable market into adjacent verticals like life sciences, robotics, and operational automation. The Hanwha LTA alone signals deep institutional validation of Ambarella’s technology, with the company gaining market share from a legacy internal SoC supplier while simultaneously creating a platform product that can be sold to other customers—turning a single partnership into a scalable revenue engine with embedded optionality for future generations of silicon.
Ambarella’s new product pipeline, particularly the 5nm CV72 and CV75 SoCs capable of running transformer-based GenAI and Agentic AI workloads, is entering steep production ramp and is expected to drive material incremental revenue in FY27, with average selling prices significantly above the current $15 corporate ASP. The upcoming 2nm CV3-AD AI SoC, slated for production in the first half of FY28, targets advanced driver-assistance systems and autonomous driving applications, leveraging Ambarella’s power-efficient architecture to address the growing semiconductor content per vehicle in automotive—where market research anticipates 10–15% annual growth despite flat to declining vehicle production. This product evolution ensures that Ambarella is not merely participating in edge AI growth but is actively shaping its trajectory through performance-leading, power-optimized silicon that commands premium pricing in high-value segments like commercial fleet telematics (where <10% of the 100M+ installed base is currently AI-enabled) and enterprise security, where operational efficiency use cases (predictive maintenance, supply chain optimization) are emerging as new growth drivers beyond traditional physical security.
The build-out of Ambarella’s indirect sales channel—including ISVs, channel partners, and system integrators—is creating a scalable go-to-market engine for fragmented markets like robotics and edge infrastructure, where technical complexity and vertical-specific expertise hinder direct sales efficiency. With over half a dozen ISVs already onboarded in verticals such as retail, healthcare, transportation, and smart cities, and plans to double this ecosystem by year-end, Ambarella is enabling turnkey solutions (e.g., real-time industrial inspection, drive-through optimization, warehouse robotics) that leverage its AI SoC software stack without requiring customers to build in-house expertise. This channel expansion reduces sales cycle friction, increases design win conversion rates, and expands the company’s addressable market beyond its traditional OEM relationships, particularly in emerging physical AI applications where end-to-end solution delivery is critical for adoption.
Ambarella's strategic positioning as one of the few companies capable of integrating all edge accelerated computing functions—CPU, AI accelerator, ISP, encoding, and system control—into a single SoC provides a durable competitive moat as AI workloads evolve toward GenAI and Agent AI at the edge. The company's CVflow architecture supports over 200 AI model architectures in production, and its unified software SDK allows seamless portability across its product stack, reducing customer integration costs and enabling faster time-to-market for complex applications like vision-language-action models in robotics and multimodal reasoning in industrial automation. This full-stack expertise is increasingly rare and valuable as customers seek to avoid fragmented multi-vendor solutions in data centers, creating a structural advantage that is difficult to replicate and positions Ambarella as a critical enabler in the physical AI and embodied AI ecosystems where real-time, low-latency processing is essential.
The execution of long-term agreements (LTAs), exemplified by the Hanwha deal with potential revenue exceeding $800 million over 10+ years, represents a transformative shift toward predictable, multigenerational revenue streams that extend beyond traditional transactional sales. These LTAs involve co-development of semi-custom SoCs, NRE funding from customers, and joint technology roadmap planning, which not only secure wafer supply commitments through partnerships with Samsung but also expand Ambarella’s addressable market into adjacent verticals like life sciences, robotics, and operational automation. The Hanwha LTA alone signals deep institutional validation of Ambarella’s technology, with the company gaining market share from a legacy internal SoC supplier while simultaneously creating a platform product that can be sold to other customers—turning a single partnership into a scalable revenue engine with embedded optionality for future generations of silicon.
Ambarella’s new product pipeline, particularly the 5nm CV72 and CV75 SoCs capable of running transformer-based GenAI and Agentic AI workloads, is entering steep production ramp and is expected to drive material incremental revenue in FY27, with average selling prices significantly above the current $15 corporate ASP. The upcoming 2nm CV3-AD AI SoC, slated for production in the first half of FY28, targets advanced driver-assistance systems and autonomous driving applications, leveraging Ambarella’s power-efficient architecture to address the growing semiconductor content per vehicle in automotive—where market research anticipates 10–15% annual growth despite flat to declining vehicle production. This product evolution ensures that Ambarella is not merely participating in edge AI growth but is actively shaping its trajectory through performance-leading, power-optimized silicon that commands premium pricing in high-value segments like commercial fleet telematics (where <10% of the 100M+ installed base is currently AI-enabled) and enterprise security, where operational efficiency use cases (predictive maintenance, supply chain optimization) are emerging as new growth drivers beyond traditional physical security.
The build-out of Ambarella’s indirect sales channel—including ISVs, channel partners, and system integrators—is creating a scalable go-to-market engine for fragmented markets like robotics and edge infrastructure, where technical complexity and vertical-specific expertise hinder direct sales efficiency. With over half a dozen ISVs already onboarded in verticals such as retail, healthcare, transportation, and smart cities, and plans to double this ecosystem by year-end, Ambarella is enabling turnkey solutions (e.g., real-time industrial inspection, drive-through optimization, warehouse robotics) that leverage its AI SoC software stack without requiring customers to build in-house expertise. This channel expansion reduces sales cycle friction, increases design win conversion rates, and expands the company’s addressable market beyond its traditional OEM relationships, particularly in emerging physical AI applications where end-to-end solution delivery is critical for adoption.
Ambarella’s financial performance remains heavily dependent on a single logistics partner, WT Microelectronics, which accounted for 60.7% of Q1 FY27 revenue, creating significant concentration risk that could disrupt operations if the relationship deteriorates or if WT shifts allegiance to competitors. Despite efforts to diversify through indirect channel expansion, the company has not demonstrated meaningful reduction in this reliance, and any disruption in WT’s fulfillment capabilities—whether due to geopolitical tensions in Taiwan, supply chain constraints, or competitive poaching—could immediately impair revenue recognition and cash conversion, especially given the company’s negative operating cash flow of $25.6 million in Q1 and rising inventory days (from 99 to 145), which suggests working capital strain tied to over-reliance on a single distribution node.
While management highlights the potential of LTAs like the Hanwha agreement, they consistently characterize the revenue as “potential” and avoid providing concrete milestones, timelines, or probability-weighted contributions to near-term guidance, raising concerns that these deals may be more aspirational than executable. The Hanwha LTA, though framed as exceeding $800 million over 10+ years, lacks clarity on upfront NRE, minimum purchase commitments, or revenue ramp rates, and history shows that semiconductor LTAs often face delays in productization, scope changes, or cancellation due to shifting customer priorities—particularly in cyclical markets like industrial automation and robotics where CapEx sensitivity is high. Without binding volume or pricing terms disclosed, investors cannot model these agreements with confidence, and the market may be overestimating their near-term impact on predictability and growth.
Ambarella’s gross margin guidance of 59–62% long-term appears vulnerable to mix shift pressures, as automotive revenue—while growing rapidly—typically carries lower gross margins than IoT due to stricter pricing pressures, longer qualification cycles, and higher customization demands in telematics and safety applications. The company acknowledged sequential double-digit growth in automotive IoT but did not disclose margin differentials between segments, and with automotive now establishing all-time quarterly records while consumer IoT declines double digits sequentially, the margin profile of the business may be deteriorating beneath the surface. Furthermore, rising inventory levels (145 days) tied to new product ramps suggest potential obsolescence risk if demand for CV7x SoCs fails to meet expectations, forcing write-downs that could erode the non-GAAP gross margin profile management aims to protect.
The company’s reliance on Samsung for leading-edge wafer supply (5nm, 4nm, 2nm) presents a critical single-point failure risk, despite management’s confidence in historical performance. While Samsung has committed to Ambarella as a 2nm customer alongside NVIDIA, the foundry is aggressively expanding its capacity for AI and HPC workloads, potentially prioritizing larger clients like NVIDIA, AMD, or its own semiconductor divisions during periods of allocation tightness. Ambarella has no long-term wafer supply contracts, leaving it vulnerable to price increases, allocation cuts, or delayed access to next-node capacity—any of which could derail product roadmaps for the CV3-AD or future nodes, especially as competitors like Qualcomm and MediaTek secure multi-year agreements with Samsung and TSMC for automotive and edge AI chips, putting Ambarella at a disadvantage in securing guaranteed access to leading-edge process technology.
Ambarella’s financial performance remains heavily dependent on a single logistics partner, WT Microelectronics, which accounted for 60.7% of Q1 FY27 revenue, creating significant concentration risk that could disrupt operations if the relationship deteriorates or if WT shifts allegiance to competitors. Despite efforts to diversify through indirect channel expansion, the company has not demonstrated meaningful reduction in this reliance, and any disruption in WT’s fulfillment capabilities—whether due to geopolitical tensions in Taiwan, supply chain constraints, or competitive poaching—could immediately impair revenue recognition and cash conversion, especially given the company’s negative operating cash flow of $25.6 million in Q1 and rising inventory days (from 99 to 145), which suggests working capital strain tied to over-reliance on a single distribution node.
While management highlights the potential of LTAs like the Hanwha agreement, they consistently characterize the revenue as “potential” and avoid providing concrete milestones, timelines, or probability-weighted contributions to near-term guidance, raising concerns that these deals may be more aspirational than executable. The Hanwha LTA, though framed as exceeding $800 million over 10+ years, lacks clarity on upfront NRE, minimum purchase commitments, or revenue ramp rates, and history shows that semiconductor LTAs often face delays in productization, scope changes, or cancellation due to shifting customer priorities—particularly in cyclical markets like industrial automation and robotics where CapEx sensitivity is high. Without binding volume or pricing terms disclosed, investors cannot model these agreements with confidence, and the market may be overestimating their near-term impact on predictability and growth.
Ambarella’s gross margin guidance of 59–62% long-term appears vulnerable to mix shift pressures, as automotive revenue—while growing rapidly—typically carries lower gross margins than IoT due to stricter pricing pressures, longer qualification cycles, and higher customization demands in telematics and safety applications. The company acknowledged sequential double-digit growth in automotive IoT but did not disclose margin differentials between segments, and with automotive now establishing all-time quarterly records while consumer IoT declines double digits sequentially, the margin profile of the business may be deteriorating beneath the surface. Furthermore, rising inventory levels (145 days) tied to new product ramps suggest potential obsolescence risk if demand for CV7x SoCs fails to meet expectations, forcing write-downs that could erode the non-GAAP gross margin profile management aims to protect.
The company’s reliance on Samsung for leading-edge wafer supply (5nm, 4nm, 2nm) presents a critical single-point failure risk, despite management’s confidence in historical performance. While Samsung has committed to Ambarella as a 2nm customer alongside NVIDIA, the foundry is aggressively expanding its capacity for AI and HPC workloads, potentially prioritizing larger clients like NVIDIA, AMD, or its own semiconductor divisions during periods of allocation tightness. Ambarella has no long-term wafer supply contracts, leaving it vulnerable to price increases, allocation cuts, or delayed access to next-node capacity—any of which could derail product roadmaps for the CV3-AD or future nodes, especially as competitors like Qualcomm and MediaTek secure multi-year agreements with Samsung and TSMC for automotive and edge AI chips, putting Ambarella at a disadvantage in securing guaranteed access to leading-edge process technology.