Ambarella : Q4 FY2025 Earnings Call Transcript
AMBA
Page 1 of 15
February 26, 2025 1:30PM PST
Q4 and fiscal year 2025 (January 31, 2025) earnings call script
Louis Gerhardy, VP Corporate Development
Good afternoon and thank you for joining our fourth quarter and full-year fiscal 2025 financial results conference call. On the call with me today is Dr. Fermi Wang, President and CEO, and John Young, CFO.
The primary purpose of today's call is to provide you with information regarding the results for our fourth quarter and full-year fiscal 2025. The discussion today and the responses to your questions will contain forward-looking statements regarding our projected financial results, financial prospects, market growth and demand for our solutions, among other things.
These statements are based on currently available information and subject to risks, uncertainties and assumptions. Should any of these risks or uncertainties materialize or should our assumptions prove to be incorrect, our actual results could differ materially from these forward-looking statements. We are under no obligation to update these statements.
1
Page 2 of 15
These risks, uncertainties and assumptions, as well as other information on potential risk factors that could affect our financial results, are more fully described in the documents that we file with the SEC.
Before starting the call, we hope to see you at one of the following investor events.
Access to our fourth quarter and full-year fiscal 2025, results press release, transcripts, historical results, SEC filings and a replay of today's call can be found on the Investor Relations page of our website. The content of today's call as well as the materials posted on our website are Ambarella's property and cannot be reproduced or transcribed without our prior written consent.
2
Page 3 of 15
Fermi will now provide a business update for the quarter, John will review the financial results and outlook and then we will be available for your questions.
Dr. Fermi Wang, President & CEO
Good afternoon and thank you for joining our call today
Ambarella finished fiscal 2025 with strong results and a positive outlook. Edge AI is clearly established as our key revenue driver, enabling us to grow and overcome the cyclical, economic and geopolitical challenges in the last year. We achieved record AI revenue in Q4 as well as for the full year fiscal 2025.
In the fourth quarter of fiscal 2025, our revenue increased 2% sequentially and exceeded the high-end of our guidance range by 5%. Our 5nm SoC products led the results, with new product revenue wave 1, from the CV5 family, leading the charge and for the first time we generated production revenue from the CV7 family, or new product revenue wave 2.
3
Page 4 of 15
Our fiscal 2025 revenue increased 26% year-over-year, with both units and average selling prices ("ASP") rising. Edge AI was about 70% of our total revenue. Our customers completed the digestion of their excess inventory in the first half of the year and in the second half the secular growth of our edge AI strategy became more apparent.
I am proud of the key achievements our Team delivered in the last year.
GAAP profitability Fiscal 2025 represented our 16th consecutive year of positive free-cash- flow.
4
Page 5 of 15
Looking into fiscal year 2026, we anticipate mid to high-teens revenue growth, despite the higher base of the much stronger than expected Q4. With uncertainty related to government policy decisions, we have built conservatism into our outlook for the second half of fiscal 2026. While we do not expect to be directly impacted by tariffs, we suspect some customers are evaluating their own supply chains as well as the elasticity of demand for their products.
Turning to developments in the broad AI market, the breakthroughs with more powerful, efficient and open-sourced reasoning models was an exciting industry development in January. We expect open-source reasoning models, such as DeepSeek's R1 to enable more advanced decision making and intelligence in all tiers of the AI processing hierarchy, including our target market at the edge. While breakthroughs like this bring new levels of compute efficiency, the implementation of the new reasoning models at the edge is expected to require incremental AI compute power. In edge AI, this is an example of a secular trend we expect to drive our average selling prices ("ASP") higher.
5
Page 6 of 15
An excellent example of how Ambarella is leading the rapid evolution of the edge AI market with innovation and execution is our CES exhibition this year where we demonstrated vision language models ("VLM") processing on our CV7, CV3 and N1 product families. In the near future we expect to demonstrate distilled DeepSeek reasoning models on our edge AI SoCs.
There is a growing trend towards performing more AI processing at the edge, as cloud based processing has a higher total cost of ownership, latency challenges, high power usage, data security and privacy concerns. Edge AI processing is now being enabled with the introduction of smaller and more optimized models.
We introduced our N1-655 edge GenAI SoC at CES this year, with high AI processing performance to support the latest transformer networks and popular multi-modal VLMs and large-language models. N1-655 consumes only 20 watts of power and targets applications such as on-premise AI boxes, autonomous mobile robots, security network video recorders ("NVRs") and retail analytics.
6
Page 7 of 15
At the show, we demonstrated the N1-655 running Contrastive Language-Image Processing
("CLIP") and Large Language Model Architecture ("LLaVA") OneVision models, with data inputs including text, video, and speech-all running locally on an AI box, without the need for an internet connection.
In the enterprise security market Motorola introduced its small form factor, V200 body worn camera, based on our S3L SoC, and designed to enhance workers safety and security in public environments such as the retail, medical or hospitality industries. Additionally, Motorola introduced its V700 body camera, based on our S6LM SoC and offering 12-hour operation.
Motorola also introduced its H6A fisheye camera based on our 5nm CV72 and offering 360- degree viewing, 11-MP resolution and AI powered video and audio analytics.
Japanese enterprise security leader iPro, formerly Panasonic, introduced its new line of High Zoom Bullet cameras targeting long range monitoring of highways, parking lots and stadiums. Based on our 5nm CV52, the cameras are pre-installed with nine edge AI applications.
7
Page 8 of 15
We are pleased to announce our first win with the HID division of Assa Abloy, a global leader in the access solutions market. HID has deployed its U.ARE.U Camera Identification System, based on our CV22, to deliver fast, accurate, and secure facial recognition. Powered by AI- driven multispectral imaging and advanced presentation attack detection, it enhances user convenience while reducing fraud risks, even in high-security and challenging environments.
In the smart home security market, Canadian communications operator Telus introduced a range of cameras including doorbell, indoor and outdoor cameras all based on our CV28.
In vehicles, a vast majority of our revenue is currently in the ADAS market, and our global effort is demonstrated with this quarter's representative customer engagements in China, South Korea and France.
During the quarter FAW introduced its Hongqi H5 PHEV including level 2 ADAS, implemented with a 1V1R ADAS system based on our CV22AQ.
VW's joint-venture with FAW introduced a pre-installed dashcam in its Magotan passenger vehicle, based on our A12AQ.
8
Page 9 of 15
NIO, considered one of the leading New Energy Vehicle companies, introduced its flagship EV, the ET9, including an electronic rear view Mirror system based on our CV28AQ.
In December, we earned our first driver monitoring win in a Hyundai-Kia passenger vehicle. The pre-installed system is provided by LG Electronics and is based on our CV25.
In January, Ford Trucks was announced as the first customer for French tier 1 Gauzy, who is
providing its AI-powered Smart-Vision® camera monitor system using our CV2FS.
In summary, this quarter's list of 10 representative customer engagements highlights the increasing breadth of our edge AI revenue globally. This includes a range of enterprise and consumer driven IoT applications, and a variety of automotive applications such as L2 ADAS, driver monitoring, rearview and exterior left and right-side mirrors.
Cumulatively, we have shipped about 30 million edge AI SoCs, with each SoC integrating our proprietary CVflow deep learning AI accelerator with our proprietary video processor. A vast majority of this installed base is represented by our CV2 family of computer vision processors, where we expect continued growth.
9
Page 10 of 15
Our new 5nm products, which command an above average ASP, represent our key growth driver, with revenue expected to arrive in waves. The first wave, from the growing CV5 family, is underway, and it continues to have a strong growth outlook. Wave 2, from the CV7 family, reached production status for the first time in the fourth quarter with 3 customers purchasing production quantities. Combined, these first 2 waves are expected to represent more than half of our incremental revenue in fiscal 2026.
In addition to the new product efforts I have described, we remain committed to develop new technologies and products capable of processing advanced AI models. As the business case develops, we will provide more information on the timing of their revenue contribution.
This is a very exciting period with the rapid evolution of the AI industry. Our edge AI business targets multiple industries where the megatrends of safety, security and automation intersect. As a result, in this dynamic environment we carefully evaluate new opportunities to maximize our return-on-investment.
10
Disclaimer
Ambarella Inc. published this content on February 27, 2025, and is solely responsible for the information contained therein. Distributed by Public, unedited and unaltered, on February 27, 2025 at 04:21:05.333.