Ambarella posts 13% revenue rise on edge AI SoC demand
Ambarella reported second-quarter fiscal 2027 revenue of $108.1 million, a 13.2% increase from $95.5 million in the same period a year earlier, as demand for its 5nm edge AI system-on-chip (SoC) products continued to expand across automotive and IoT end markets. The Santa Clara-based semiconductor company also narrowed its GAAP net loss sharply, from $20.0 million in Q2 FY2026 to $6.7 million this quarter, partly assisted by a one-off $9.0 million credit arising from the termination of a development project.
On a non-GAAP basis, which strips out stock-based compensation, acquisition-related costs and the one-off project credit, Ambarella posted net income of $8.2 million, or $0.18 per diluted share, up from $6.4 million and $0.15 a year ago. Non-GAAP gross margin was 59.3%, a modest decline from 60.5% in the prior-year quarter. The company held $272.3 million in cash, equivalents and marketable debt securities at the period end, slightly down from $277.8 million at the close of the prior quarter but ahead of $261.2 million a year ago.
Strategic moves and new products
President and chief executive Fermi Wang said edge AI revenue reached record levels in the quarter, with growth balanced across the automotive and IoT verticals. Wang highlighted two distribution agreements, each running seven years, with Japanese technical distributor Macnica and consulting and engineering firm Capgemini, describing them as part of a broader effort to build out an indirect sales channel. Ambarella also introduced what it is calling its first stand-alone AI accelerator, the X7, which the company says is intended to extend its addressable market into higher-value edge inference workloads.
Wang stated that these developments are contributing to an upward revision in the company's five-year serviceable addressable market forecast for edge AI and Physical AI, though Ambarella did not publish specific revised SAM figures alongside the results.
Market context and competitive landscape
Ambarella occupies a specialised niche in the edge AI semiconductor market, competing with a broader set of chip vendors as AI inference workloads migrate from centralised data centres toward endpoint devices. Rivals include companies such as Qualcomm and NXP Semiconductors in automotive-grade silicon, as well as a cluster of startups targeting vision-based edge inference. Nvidia's Jetson platform also competes in the industrial and robotics segments.
The company's installed base of more than 50 million AI SoC units gives it a meaningful deployment footprint, particularly in physical security cameras, advanced driver-assistance systems, and drone applications. The shift toward Physical AI, combining perception, decision-making and control at the device level, is opening new categories beyond the company's traditional video-processing heritage, and the seven-year Capgemini agreement suggests Ambarella is betting on systems integration partners to reach enterprise buyers it cannot address through direct sales alone.
Export-control dynamics remain a background risk. US Bureau of Industry and Security rules on advanced semiconductor exports continue to evolve, and edge AI SoCs with high-performance neural network accelerators may face additional scrutiny depending on end-use and geography, particularly as the automotive AI supply chain increasingly spans US, European and Asian manufacturers.
Outlook
For the third quarter of fiscal 2027, ending 31 October 2026, Ambarella guided revenue of $115.0 million to $124.0 million, implying further sequential growth. Non-GAAP gross margin is expected in the range of 59.0% to 60.0%, with non-GAAP operating expenses between $56.5 million and $59.5 million. Investors will be watching for early commercial traction from the X7 accelerator, evidence that the Macnica and Capgemini channel agreements are generating design wins, and any update on the revised SAM estimate Wang referenced.