0001193125-26-119321
SEC filingRevenue surged 37.2% to $390.7M on AI processor demand, but gross margin slipped, and operating loss narrowed.
Ambarella is a leading developer of low-power system-on-a-chip (SoC) semiconductors and software for edge and physical AI applications and intelligent automation. The company’s technologies enable electronic systems to become partially or fully autonomous with features such as intelligent automation, complex scene understanding, and autonomous decision-making. Until 2023, a majority of revenue came from human-viewing applications, but recent development efforts have focused on advanced AI inference that enables edge devices to perceive and interact with the physical world using data from cameras and other sensors like 4D radar. This shift targets the physical AI subset of the edge AI market.
The company addresses two broad end markets: IoT Applications and Automotive Applications. IoT includes security cameras (enterprise, home, public) and portable video applications (body-worn cameras, sports/action cameras). Automotive includes ADAS/safety (front ADAS cameras, cabin monitoring, AI telematics, electronic mirrors, video recorders) and autonomy (central domain controllers for L2+ to L4 autonomous vehicles and emerging robotic applications). Revenue share by segment is not disclosed.
Key product families include the CV7 (CV7, CV72, CV75) for high-performance edge AI, the CV3 family (e.g., CV3-AD685) for automated driving and robotics, and the N1 SoC for edge generative AI supporting models up to 34 billion parameters. Additional products include Vision Processor SoCs, High Definition Radar (adaptive AI software), SerDes (B8), and software modules (NN-ISP, ADAS stack, etc.). The Cooper Development Platform integrates hardware, software, and AI models to accelerate time-to-market.
Ambarella sells directly and through distributors to ODMs, OEMs, and Tier-1 automotive suppliers. The sales cycle involves design wins leading to volume production in 12-18 months. Sales are primarily via purchase orders. Asia accounted for 88% of fiscal 2026 revenue, with WT Microelectronics representing 70% of total revenue. Customer concentration is high, and the company relies on a limited number of customers for a significant portion of revenue.
The semiconductor market is highly competitive. In IoT, primary competitors include HiSilicon, Novatek, NVIDIA, Qualcomm, and Sigmastar. In automotive, competition includes Horizon Robotics, Mobileye, Novatek, NVIDIA, Qualcomm, Renesas, and Texas Instruments. Many competitors have greater financial and technical resources. Ambarella competes on AI performance, power efficiency, image quality, software ecosystem, and total solution cost.
Ambarella’s strategy focuses on targeting IoT and automotive markets requiring advanced AI inference. The company leverages its proprietary CVflow architecture for efficient processing of transformer networks and multi-modal data fusion (cameras, radar, lidar, etc.). It aims to provide scalable, low-power SoC platforms with comprehensive software, and drives ecosystem adoption through the Cooper Developer Platform and developer zone offering optimized models and agentic blueprints. Ongoing R&D emphasizes advanced edge AI and physical AI capabilities.
As of January 31, 2026, Ambarella employed 959 people globally: 252 in the U.S., 612 in Asia (372 in Taiwan, 225 in China), and 95 in Europe. Approximately 75% of employees are in R&D, 23% in sales/marketing/admin, and 2% in operations. The voluntary turnover rate was 7.7% for fiscal 2026. The company emphasizes attracting and retaining talent through competitive compensation, equity incentives, and benefits tailored to local markets.
For fiscal year 2026, Ambarella reported revenue of $390.7 million, a 37.2% increase from $284.9 million in fiscal 2025. The growth was driven by higher product unit shipments and an increased percentage of sales from higher average selling price AI inference processors, reflecting strong demand for edge AI solutions. This was partially offset by lower nonrecurring engineering (NRE) project service revenue. Gross margin declined to 59.2% from 60.5%, primarily due to higher manufacturing costs associated with advanced process technologies and lower sales of previously reserved inventory, though partially offset by the favorable mix shift to AI processors. Operating loss improved significantly to $82.5 million from $126.6 million, as the increase in gross profit ($58.9 million) outpaced the rise in operating expenses ($14.9 million). Operating expenses grew 5.0%, driven by higher R&D costs (engineering-related, personnel, and facility relocation) and SG&A increases (personnel and marketing). Net loss narrowed to $75.9 million from $117.1 million, with income tax expense shifting from a benefit to a provision due to reduced pre-tax loss and lower tax reserve releases.
The MD&A does not provide discrete segment-level revenue or profit. Historically, Ambarella’s revenue is concentrated in IoT and automotive end markets, which collectively generated the majority of revenue. Management notes that AI inference processors, which carry higher average selling prices, are becoming a larger share of sales, driving the mix shift. The company continues to target emerging applications such as AI-enabled security cameras, ADAS, and robotics, but does not break out performance by end market or product category in this section.
Management expects R&D expenditures to increase in absolute dollars as the company invests in innovative video and image processing solutions with AI capabilities and targets new markets. They anticipate that future revenue growth will depend on expanding within camera markets with AI technology, particularly IoT, and emerging markets like AI-based driving applications and industrial robotics. The company believes its existing cash balances ($312.6 million as of year-end) will be sufficient for at least the next 12 months. No specific revenue or margin guidance is provided. The outlook focuses on design wins and product life cycles, with typical IoT product cycles of 12–24 months and longer cycles in automotive and industrial markets. Management also notes ongoing global supply chain uncertainties as a factor to monitor.
As of January 31, 2026, Ambarella held $191.0M in cash and cash equivalents and $121.6M in marketable debt securities, totaling $312.6M in liquid assets. Total shareholders' equity was $594.8M. The company has no debt, with only operating lease liabilities of $13.4M (undiscounted payments $18.2M). Inventory increased to $52.2M from $34.4M a year earlier, reflecting work-in-progress ($35.5M) and finished goods ($16.7M). Deferred revenue (current) stood at $22.4M, while total unsatisfied performance obligations (RPO) were $76.1M, 97% expected within 12 months.
Manufacturing purchase commitments totaled $80.4M as of January 31, 2026, up from $56.4M in the prior year, driven by sales forecasts. These commitments are cancelable but may incur costs. Additionally, other noncancellable service contracts amounted to $12.4M. No loss liabilities were recorded for adverse purchase commitments. The company also has software license liabilities of $24.2M ($11.4M current, $12.8M long-term).
During FY2026, Ambarella repurchased 24,152 shares for $1.0M. As of January 31, 2026, $48.0M remained under the share repurchase program, which was extended through June 30, 2026. Subsequent to year-end, an additional 18,702 shares were repurchased for $1.0M, leaving $47.0M available. No dividends were paid, and the company has no debt. Capital expenditures (purchases of tangible and intangible assets) were $15.5M, representing 4.0% of revenue, primarily for software licenses and property and equipment.
Ambarella operates as a single reportable segment. Revenue by geography (based on bill-to location) shows heavy concentration in Taiwan ($271.9M, 70% of total), followed by Asia Pacific other than Taiwan ($71.0M, 18%), Europe ($20.1M), North America excluding US ($21.8M), and United States ($5.8M). The customer WT Microelectronics accounted for 70% of revenue, with $24.6M in accounts receivable. The company's property and equipment are located primarily in the United States ($6.6M), Taiwan ($3.6M), Europe ($1.1M), and other Asia Pacific ($0.3M).
Ambarella faces significant regulatory risks from US-China trade tensions. Export controls now restrict sales to certain Chinese customers on the BIS Entity List, and further restrictions could materially impact revenue. The company also highlights risks from potential Chinese retaliation and increased tariffs. Geopolitical instability in Taiwan (where 372 employees are based) and Hong Kong (primary warehouse) could disrupt operations.
The company relies heavily on Samsung for wafer fabrication, especially for advanced nodes (5nm and below). Past supply constraints and the lack of long-term contracts create vulnerability. Single-sourcing at each foundry means any disruption (e.g., natural disasters in Taiwan or Korea) could halt production. Additionally, the primary warehouse in Hong Kong is exposed to political and economic risks.
Ambarella competes with much larger players like NVIDIA, Mobileye, and HiSilicon. The lengthy design win cycle (12-18 months for IoT, longer for automotive) requires significant upfront investment without guaranteed returns. Customer concentration is extreme: distributor WT accounts for 70% of revenue, and top 10 end customers represent ~67%. Loss of a major customer (e.g., Insta360) would be severe. The AI market is still nascent, and uncertain adoption or regulatory backlash could limit growth.
The reliance on a single distributor and a few end customers poses a high risk. The company’s operating results are volatile due to cyclical semiconductor demand, order cancellations, and macroeconomic factors. Average selling prices tend to decline over time, pressuring margins if costs cannot be reduced.
Rapid technological change requires continuous R&D investment ($238.5M in FY2026). Transitioning to smaller process nodes (e.g., 2nm) is costly and dependent on foundries. Defects or bugs in complex SoCs could harm reputation. Also, the adoption of open-source software may risk exposure of proprietary code.
The semiconductor industry is litigious. Ambarella may face IP infringement claims, which could lead to costly litigation or invalidate patents. Indemnification obligations to customers could amplify liabilities. The company also must comply with evolving data privacy laws (GDPR, CCPA, China PIPL) and anti-corruption regulations (FCPA), which add compliance costs and risk of penalties.
No cash flow data available in the provided text.