Ambarella reported stronger fiscal second-quarter 2027 revenue and a substantially narrower quarterly loss as the semiconductor designer benefited from expanding demand for artificial intelligence processing at the edge, reinforcing management’s argument that the company is moving beyond its traditional video-processing markets toward a broader physical AI platform. Revenue for the quarter ended July 31 reached $108.1 million, an increase of 13.2% from $95.5 million in the year-earlier period. The result also came in slightly above the roughly $107.8 million expected by analysts tracked ahead of the report.

The quarter marked another period of double-digit growth for the Santa Clara, California-based chipmaker. Ambarella had reported first-quarter fiscal 2027 revenue of $100.4 million, making the second-quarter result an increase of roughly 8% sequentially as well as a double-digit gain from a year earlier. For the first six months of fiscal 2027, revenue totaled $208.5 million, up 14.9% from $181.4 million in the comparable period of fiscal 2026. The progression gives the company a stronger revenue base entering the second half of its fiscal year.

Profitability improved sharply on a GAAP basis. Ambarella recorded a second-quarter net loss of $6.7 million, equivalent to $0.15 per diluted ordinary share, compared with a loss of $20.0 million, or $0.47 per share, a year earlier. Its operating loss narrowed to about $8.1 million from approximately $22.0 million. Gross profit increased to $62.4 million from $56.2 million as higher revenue more than offset the decline in gross-margin percentage.

The headline reduction in the GAAP loss requires an important qualification. Ambarella recognized a one-time $9.0 million reduction in research-and-development expense after releasing a deposit liability associated with the termination of a development project. The company excluded that credit from its non-GAAP calculations because management considered it non-recurring. Research-and-development expense consequently fell to $50.6 million on a GAAP basis from $59.7 million a year earlier, while total GAAP operating expenses declined to $70.6 million from $78.2 million. Chief Financial Officer John Young later said the terminated program involved an automotive autonomy customer and was not one of the semi-custom projects Ambarella has been discussing as part of its future growth strategy.

On the company’s adjusted measures, improvement was more moderate but still positive. Non-GAAP net income rose to $8.2 million, or $0.18 per diluted share, from $6.4 million, or $0.15 per share, in the year-earlier quarter. For the first half of the fiscal year, non-GAAP net income totaled $13.3 million, or $0.30 per diluted share, compared with $9.5 million, or $0.22 per share, a year earlier. Ambarella’s adjusted earnings per share was also slightly ahead of the roughly $0.17 consensus cited around the earnings release.

Margins were less favorable. GAAP gross margin declined to 57.7% from 58.9% a year earlier, while non-GAAP gross margin fell to 59.3% from 60.5%. The direction is notable because Ambarella’s growth strategy increasingly relies on advanced process technologies and higher-performance artificial intelligence processors, creating a balance between higher average selling prices and the manufacturing costs associated with more sophisticated silicon. In its fiscal 2026 annual filing, the company had already identified higher manufacturing costs for advanced process technologies as a factor pressuring gross margin, partly offset by a richer mix of higher-priced AI inference processors.

Management nevertheless presented demand as healthy. President and Chief Executive Officer Fermi Wang said edge AI revenue reached a record during the quarter, with sequential growth in both automotive and IoT markets. He highlighted particularly strong growth from Ambarella’s CV75 and CV72 systems-on-chip, both produced on 5-nanometer technology. Automotive revenue also reached a company record, according to management’s earnings-call commentary, supported by continued adoption of AI processing in commercial vehicles.

Ambarella’s strategy centers on performing artificial intelligence inference close to where data is created rather than sending every workload to centralized cloud infrastructure. Its chips are used in applications including vehicle safety systems, security cameras, telematics, industrial equipment, drones and robotics. The company says more than 50 million of its AI systems-on-chip have been deployed. Edge inference can be particularly valuable where latency, power consumption, privacy, communications bandwidth or operating reliability makes continuous cloud processing impractical.

The next-quarter outlook indicates that management expects the current growth momentum to continue. Ambarella forecast fiscal third-quarter revenue of $115 million to $124 million for the period ending October 31. The midpoint of $119.5 million would represent an increase of approximately 10.5% from the second quarter. The company expects non-GAAP gross margin of 59% to 60% and non-GAAP operating expenses between $56.5 million and $59.5 million.

That forecast keeps the focus on revenue expansion rather than an immediate margin breakout. The midpoint of the third-quarter gross-margin range is close to the second-quarter adjusted result, and management continues to cite a longer-term non-GAAP gross-margin objective of 59% to 62%. On the earnings call, executives said the new indirect distribution strategy involving Capgemini and Macnica did not currently require the company to revise that long-term margin framework.

A potential complication for the second half is the global memory market. Management described memory availability and pricing as an industry-wide concern as suppliers prioritize demand from AI data centers. Ambarella does not buy and resell the memory incorporated alongside its processors, meaning rising DRAM prices do not directly reduce its semiconductor gross margin. The risk is indirect: customers facing higher memory bills may raise product prices, cut production or reduce chip orders. Management said it had seen little revenue impact from the memory situation in the fiscal second and third quarters but was continuing to monitor potential effects on the fourth quarter.

Ambarella is also attempting to turn power efficiency into a competitive advantage in that environment. Management argues that its processors can handle edge AI workloads with lower power and memory requirements than more general-purpose GPU-based alternatives. That positioning is central to the company’s recently introduced X7, its first stand-alone AI accelerator. Unlike an integrated Ambarella system-on-chip, X7 can be paired with a separate host processor, including Ambarella devices or other CPU platforms, allowing customers to add additional AI performance without redesigning an entire computing board.

The X7 illustrates a broader shift in Ambarella’s product strategy. The company historically built highly integrated processors combining video, image signal processing and artificial intelligence functions. It is now expanding toward stand-alone accelerators, edge infrastructure systems and semi-custom designs that potentially address a larger range of enterprise and industrial computing deployments. Management sees growing opportunities in robotics and other forms of physical AI, where machines combine perception, inference and decision-making in real time.

That shift is reflected in Ambarella’s updated assessment of its addressable opportunity. On the earnings call, management said its rolling five-year serviceable available market forecast now rises from approximately $8.5 billion in fiscal 2027 to $22.9 billion in fiscal 2032, representing a compound annual growth rate of about 20%. IoT-related markets are expected to represent roughly 70% of the terminal-year opportunity, with edge infrastructure emerging as an important contributor. The forecast is a management estimate rather than contracted revenue, but it shows how much of Ambarella’s investment case now rests on applications beyond its established camera and automotive businesses.

The company paired its quarterly report with two long-term commercial initiatives intended to reach those markets. Ambarella announced a strategic engagement with Capgemini to help enterprises develop and deploy edge and physical AI systems in areas including smart infrastructure, retail, logistics, industrial automation, healthcare and automotive applications. Capgemini is expected to contribute engineering, systems-integration and industry expertise, reducing the implementation burden for customers that may have limited experience deploying specialized AI hardware outside centralized data centers.

Ambarella separately entered a seven-year strategic distribution agreement with Macnica covering the Americas and Europe, the Middle East and Africa. Macnica will distribute Ambarella’s edge AI infrastructure portfolio, including CV-series and N-series processors, systems-on-module products, development platforms, reference designs and software. The arrangement is aimed in part at independent software vendors, original equipment manufacturers, original design manufacturers and systems integrators that Ambarella’s direct sales organization has not historically served at scale.

Management cautioned that the economic contribution from those relationships will take time. During the earnings call, Wang said initial design wins could begin generating modest revenue as early as next year, but meaningful revenue contributions that materially affect Ambarella’s forecasts could require two to three years. Management characterized each relationship as potentially representing roughly $500 million of opportunity over its lifetime, though such estimates depend on future design wins, customer adoption and production ramps rather than firm near-term sales commitments.

For investors evaluating the current quarter, the more immediate indicators remain the expansion of core edge AI revenue and the path toward sustainable profitability. Ambarella ended the second quarter with $272.3 million in cash, cash equivalents and marketable debt securities, compared with $277.8 million in the preceding quarter and $261.2 million a year earlier. The liquidity position gives the company capacity to continue investing heavily in semiconductor design and software while its newer products and distribution initiatives move through lengthy customer qualification cycles.

Stock-based compensation remains a meaningful difference between Ambarella’s GAAP and adjusted results. The company recorded $22.7 million of stock-based compensation in the second quarter, down from $25.2 million a year earlier. Together with acquisition-related expenses and the treatment of the development-project credit, those adjustments explain much of the gap between the reported GAAP loss and the $8.2 million of non-GAAP income. Investors therefore have several profitability measures to assess: a sharply narrower statutory loss, positive adjusted earnings, and a gross-margin profile that remains below the year-earlier level.

The second-quarter numbers ultimately strengthen Ambarella’s growth narrative without eliminating execution risks. Double-digit revenue expansion, record edge AI sales and another sequentially higher revenue forecast suggest that newer AI processors are gaining traction. At the same time, the one-time R&D credit amplified the improvement in GAAP profitability, while gross-margin compression shows that revenue growth is not yet translating cleanly into stronger unit economics.

The central earnings question for the coming quarters will be whether Ambarella can maintain double-digit growth while protecting its targeted margin range as customers navigate higher component costs. Progress in automotive, enterprise edge infrastructure and robotics would broaden the company’s revenue base and increase exposure to higher-value AI workloads. Conversely, weaker customer production caused by memory constraints or slower adoption of new edge AI platforms could delay the operating leverage investors are seeking.

For fiscal third quarter 2027, Ambarella has set a measurable near-term test: revenue of at least $115 million, with the top of the range reaching $124 million, alongside adjusted gross margin of 59% to 60%. Delivering within that framework would extend the company’s revenue recovery and provide further evidence that edge AI demand is moving from design activity into commercial shipments. The longer-term valuation case, however, will depend on whether Ambarella can convert its expanding product portfolio and new distribution channels into durable revenue growth without sacrificing the profitability gains that have begun to emerge in fiscal 2027.