管理層發言
Good morning, and welcome to the Ambiq Micro Second Quarter 2026 Earnings Conference Call. As a reminder, this conference call is being recorded. I'd now like to turn the call over to Ms. Charlene Wan, Ambiq's Vice President of Corporate Marketing and Investor Relations. Charlene, please go ahead.
On today's call, Ambiq's CEO, Fumihide Esaka, will provide an overview of the company's performance and strategy. CFO, Jeffrey Winzeler, will then discuss the quarter's financial results and outlook. Following their remarks, Scott Hanson, Ambiq's Founder and CTO, will join Fumi and Jeff for Q&A. Our earnings release is available on the Investor Relations page of our website at www.ambiq.com. We have also posted our earnings presentation on the Investor Relations section of our website. Before I turn the call over to Fumi, I'd like to remind our listeners that during the course of this conference call, management will discuss non-GAAP financial measures. Reconciliations of these non-GAAP measures to their most directly comparable GAAP measures are included in our earnings release available on the company's Investor Relations website. In addition, today's call will contain forward-looking statements representing management's beliefs and assumptions only as of the date made. Our most recent quarterly report on Form 10-Q and other filings with the SEC provide more information on specific risks that may cause the actual results to differ materially from current expectations. And now it's my pleasure to turn the call over to Ambiq's CEO, Fumi Esaka.
Good morning, everyone, and thank you for joining us. Since the start of the year, we've seen a step change in demand for Edge AI, which supports our strong performance and further increases our conviction in the magnitude and durability of the long-term opportunity. Across end markets, companies are embedding more sophisticated AI into a broader range of devices and end-user demand is far exceeding our expectations and those of our customers. With our full stack ultra-low power solutions, Ambiq's defining technologies are not only enabling but further accelerating this next chapter in AI. This positions us to continue outpacing the broader market as we take share and expand our addressable opportunity. This momentum is reflected in our second quarter results. Net sales were ahead of guidance, growing approximately 90% over last year and marking our fifth consecutive quarter of sequential growth. We also raised approximately $168 million in net proceeds through a successful follow-on offering. The strong investor interest serves as a proof point of confidence in our ability to capitalize on the meaningful opportunity ahead. Turning to the details of the quarter. Demand accelerated across customers, end markets and products. Customer programs grew on healthy end-user demand, positive response to customers' recent launches and continued ramping of our newest large customer. This supported strong performance across key products, including double-digit growth in Apollo II and Apollo IV, while Apollo V sales more than doubled year-over-year. With lean inventory across channels and continued expedite requests, we are confident that this demand reflects underlying end market consumption rather than inventory replenishment. Orders continue to exceed initial forecasts. And based on our ongoing customer conversations, we expect underlying demand for Edge AI to strengthen even further in the second half of 2026. At the same time, we and the broader semiconductor industry are increasingly supply constrained. We are working closely with our supply chain partners to support production ramps while maintaining the high level of quality and execution our customers expect. With tight capacity across wafers, packaging, substrates and testing, we are actively working to secure additional foundry allocations and OSAT capacity to improve supply flexibility and better support customer demand over time. Even with these actions, our second half outlook is constrained by available supply given the significant levels of demand we are seeing. As additional capacity becomes available, we believe we are well positioned to convert this deferred demand into future revenue. Based on our current visibility, we expect third quarter net sales to grow approximately 100% year-over-year, and we now expect to deliver approximately $135 million in net sales for the full year, even with the supply constraints I just mentioned. Looking ahead, we believe the opportunity for Ambiq is substantial. With more than 20 billion connected devices already deployed, there is significant potential to embed Edge AI across this large and growing installed base. While adoption remains at an early stage, increasing AI functionality at the edge is driving demand for greater compute and memory content across a growing range of devices and use cases. As these workloads become more sophisticated, customers must deliver higher levels of intelligence and performance without compromising battery life, responsiveness or privacy. That challenge is making power efficiency a critical design requirement for the next generation of edge devices. These requirements are redefining what matters most in semiconductor design. Power efficiency is no longer just an advantage. It is becoming a fundamental requirement for the next generation of edge AI devices. That is exactly where Ambiq is uniquely positioned to lead and create value. To capture this opportunity, we are broadening our reach across form factors, customers and end markets while investing in the next generation of products that will push the boundaries of ultra-low power performance. Our latest hardware and software innovations are further enhancing our ability to support increasingly advanced AI workloads at the edge. On the hardware side, we recently introduced the Apollo 330 Plus and Apollo 510 Light SoC families. Both products are generating strong customer demand with backlog tied to next-generation product road maps. We expect these products to begin contributing revenue in the third quarter with the first customer devices expected to reach the market early next year. We are also expanding the value of our solutions through a growing suite of AI software capabilities. Our recently launched HeliaCORE and compression kit are already helping customers accelerate production deployments, improve power efficiency, reduce memory requirements and lower overall system costs. More recently, we introduced Helia Profiler, an open source profiling tool that broadens the Helia AI ecosystem and gives developers greater visibility into model performance and system optimization. The breadth and depth of our full stack portfolio is enabling us to support more applications, win new customers and expand into diverse end markets. This includes greater penetration in medical, industrial and smart home and building applications. We continue to expect revenue from these markets to more than double in 2026, making them a more meaningful contributor to Ambiq's growth and diversification. And we expect this momentum to continue, supported by a growing and increasingly diverse design funnel. For example, in wearable neurotechnology, on-device AI is being used to analyze complex biosignals like EEG in real time, improving responsiveness and privacy. In smart buildings, leak detection systems can distinguish meaningful events directly at the sensor, reducing data transmission while enabling faster intervention. And in industrial applications, always-on sensors monitor equipment health in real time, detecting failures before they disrupt operations. These are exactly the types of workloads our ultra-low power SoCs and innovative AI software are designed to support, positioning us well to capture the next wave of edge AI adoption across multiple high-growth markets. Turning to our product road map. We continue to advance our next-generation products, Apollo 340, Atomiq 110 and Atomiq 120. Customer interest remains incredibly strong, including engagement with Alpha customers as these products are critical enablers for our customers to advance their own ambitious AI road maps. Development for these products continues to advance with both Atomiq 110 and Apollo 340 targeted for customer sampling in early 2027. In closing, the customer demand we are seeing today reinforces our belief that Edge AI is becoming one of the most important long-term growth drivers in the semiconductor industry. While near-term supply constraints remain an industry-wide headwind, we are taking decisive actions to support our customers. As we look ahead, we are exceptionally well positioned to capture the meaningful long-term opportunity in Edge AI through differentiated technology and expanding product portfolio and growing customer engagement. With that, I will turn it over to Jeff to cover the financials.
Thank you, Fumi, and good morning, everyone. We delivered a strong second quarter with non-GAAP gross profit more than doubling year-over-year. This marks our fifth consecutive quarter of growth and reflects the strength of our differentiated technology, accelerating Edge AI demand environment and strategic shift to higher-value market opportunities for our products. Now turning to the details of our second quarter financial results. Net sales were $33.9 million, increasing 89.7% year-over-year with revenue outside of our 3 largest customers growing 143% year-over-year. Sales to end customers in China were approximately 14% of total net sales compared to approximately 12% in the prior year period. Our continued strategy with regard to China is to engage in customer programs where our technology is enabling higher-value edge AI functionality. Non-GAAP gross profit increased 109.3% year-over-year to $16 million. Non-GAAP gross margin was 47.2%, up 450 basis points year-over-year on favorable mix related to greater edge AI enablement as well as improved manufacturing efficiencies. Turning to operating expense. Non-GAAP R&D was $11.2 million, up 55.5% year-over-year as we increased investments in product development and technology. The primary drivers for R&D increases were intellectual property licensing for multiple product developments, compensation costs as we scale the team and contractor costs to augment our own labor. Non-GAAP SG&A expenses were $8.2 million, up 23.7% year-over-year, driven largely by sales compensation for higher revenues and public company costs. Second quarter non-GAAP net loss was $1.8 million, a $4.1 million improvement year-over-year. Non-GAAP net loss per share was $0.07 based on 21.74 million average shares outstanding. We ended the quarter with no debt and $366.8 million in cash and cash equivalents, which included approximately $168 million in net proceeds from our upsized follow-on offering in June of this year. In total, our 2 offerings this year raised approximately $243 million in net proceeds, providing us the financial flexibility to fund working capital, sales and marketing and product development as we continue to capitalize on the growing edge AI demand and scale to meet customer needs. Now turning to our outlook. For the third quarter, we expect net sales in the range of $36 million to $37 million, driven by the trends covered by Fumi earlier. We expect non-GAAP gross margin between 46.5% and 47.5%, which is consistent with the second quarter performance. Non-GAAP operating expense of $24 million to $25 million, reflecting investments to support product development and strategic growth priorities, including $2 million related to intellectual property purchases in the quarter. Finally, we expect a non-GAAP loss per share of $0.20 to $0.12 based on a weighted average share count of 24.17 million shares outstanding. This new share count is reflective of our follow-on offering in June. Looking ahead, customer demand is accelerating meaningfully as we enter the second half of the year. At the same time, we are navigating increased supply constraints and rising cost pressures alongside the broader industry. Despite this, we are on track to double year-over-year net sales growth in the second half, positioning us to deliver approximately $135 million for the full year. For gross margin, we now expect modest year-over-year improvement in 2026 compared to our prior 2026 expectations for flat margins year-over-year. And we expect to achieve this even with the broader industry headwinds we are navigating. We continue to expect operating expense of approximately $85 million for the full year, including $7 million to $10 million of IP purchases necessary for product development. With that, I'll turn the call back over to Fumi before we open the line for Q&A.
We are pleased with our performance in the first half of 2026 and encouraged by the momentum we continue to see across the business. We remain focused on delivering for our customers, expanding capacity, advancing our product road map and investing for long-term sustainable growth. We believe the foundation we have built positions Ambiq well for the opportunities ahead, and we remain confident in our ability to execute. With that, I will open the call to questions. Operator, please go ahead.
分析師問答
Your first question comes from the line of Tore Svanberg with Stifel.
Fumi, Jeff, congratulations on the strong results. Fumi, I was hoping you could elaborate a little bit more on the supply constraints. I guess the question is, had you had the capacity, how much more could 2026 be? And when do you expect some of these supply issues to ease?
Tore, thanks for the great question. Our customers' demand is skyrocketing as we said in our statement. And as we speak, our demand keeps going up. So I cannot put a specific number, but I must say that a lot of additional demand is coming in week after week. And we believe that this trend will continue not only in the second half of this year, but even into 2027. That said, because of the fact that our demand continues to grow faster than the market, I believe that the supply constraint is something that we need to diligently work on with our supply chain partners. But as you can see, we do have a very strong partnership with our supply chain partners. So despite the fact that we are doubling our revenue year-over-year, we have been able to meet that demand, and we believe that we will be able to meet forecasted quantity as closely as possible. Again, one challenge is that the customer success has been phenomenal. Let me tell you one example. A couple of customers introduced a brand-new product back in May. Their preorder quantity was three to five times what even they expected. So to meet that demand, they wanted product in June, which was physically impossible. So we couldn't do that. But we will work with our supply chain partners and the customer to sustain our continuous very strong growth.
Yes. No, that's great color. And maybe as my follow-up question for you, Scott, and specifically on product development and Atomiq — it sounds like the first Atomiq product is going to be sampling first half of next year. Just curious, are we still looking at a 2028 revenue ramp from the Atomiq products?
Yes. Thank you. Yes, that's all still the plan, 2028 meaningful ramp for Atomiq 110. Great progress in the last several months since our last call. Probably the most notable thing is that the early development platform is in customer hands in the form of an FPGA, and we're getting a lot of useful feedback. And there are — the sales team is building a nice list of customers that want access to that product. And so I'm excited about that. And its interest spans a couple of different markets. So that's very positive. So yes, I'm pleased with how that's going. And I will say it's a busy time for the development team because we're developing multiple products in parallel. We've got Atomiq 110. We've got Apollo 340. So a lot going on, and I'm pleased with how things are proceeding.
Your next question comes from the line of Quinn Bolton with Needham & Company.
This is Shan on for Quinn. Congrats on all the progress. I guess on the gross margin for Q2 and Q3 coming in much better than expected. So I just wanted to hear the puts and takes on what's driving the strength here, especially with the rising component costs and supply constraints.
Yes. So there's two basic things that we've really made significant progress on that have allowed us to achieve gross margins a little bit above our business model. On the top side, from an ASP perspective, we continue to win business and price our products to extract the maximum value that we're providing to end customers. So pricing continues to be something that we look at and make sure that we're balancing to get the most that we can from the products that we sell. We've also made a lot of progress on the actual manufacturing costs associated with our products. And this is primarily through yield improvements and test time improvements as we ramp products into full-scale manufacturing. That is being muted a little bit by some of these cost pressures that we're seeing, but making progress on both the top side as well as the cost basis for our products is what's allowed us to deliver these margin results.
Got it. That's helpful. And then in terms of just the strong demand in the wearables market, it sounds like it's pretty broad-based, but is there any form factor that you guys are seeing more demand for, whether that's watches, bands, rings or glasses?
Yes. We are indeed seeing demand across all types of wearables, right? So whether it's wrist-based in watches and bands, displayless bands or smart rings or even glasses. What I will say is that one of the hot new areas is displayless trackers. So that would encompass both your smart rings as well as trackers like the Loop device and like the new Fitbit Air. There's great demand for that. And I would say that what's driving that is that there's this movement towards AI agents in the cloud, gathering us all this data. You don't necessarily need the display right there to tell you what's going on. And so you rely on the AI agent to analyze all your data and give you feedback about how to adjust your sleep and how to adjust your eating and so forth. And as we've talked about in the past, these devices become almost like medical devices. So we're really excited about the future there. We see our customers being very excited about the future there. So I expect good things out of that segment moving forward.
Your next question comes from the line of Liam Pharr with Bank of America.
I was wondering if you could start with just discussing kind of your revenue mix across end markets. And especially in your funnel, is it still the markets outside of wearables comprising around 25%? Or has that picked up over the last couple of quarters?
Well, one of the edge AI devices — what we call personal devices — growth is phenomenal. So we believe that will continue to grow. That said, the nonwearable market is also growing. Like we said in the script, we said we doubled nonwearable demand. However, because the denominator is growing so fast in personal devices with Edge AI capability, that absolute number may not be as great as what we'd like to see, but its growth is very strong.
And then what I'll add is that it does remain true that if we look at new designs launching next year, roughly 25% plus of that funnel is nonwearable devices. So that remains true, but we're, to some extent, victims of our own success because everything else is going so well. So I would say good progress, but maybe not as fast as one might like in these fast-growing times.
And then $1.8 million of net loss for the quarter. I was wondering if you could just kind of provide some color on that trajectory towards profitability on the bottom line. And is Atomiq needed to get there? Or can we see some profitability or even free cash flow breakeven in fiscal '27?
Yes. I think we're very, very pleased with our progress in terms of reducing the operating loss that you're seeing quarter-over-quarter. This is a very fast-growing revenue environment. We're spinning off an even higher percentage of gross profit dollars and our spending is not growing as fast. Although what I would point to is if you look at our guidance for Q3, we expect OpEx to jump back up. And again, this is really investments around our 110 and 340 development. I think it's too early to really talk about that tipping point of when we'll get to cash flow breakeven. We're very much focused on a growth model right now. And in fact, we've raised quite a bit of money in the last two offerings that we've done specifically to give us the cash resources to both grow our existing pipeline of business as well as develop new business opportunities. And so we'll continue to be investing those dollars pretty heavily. And it's too early really to talk about when we expect to get to a profitability metric.
Your next question comes from the line of Sujeeva De Silva with ROTH Capital.
Scott, Jeff, congratulations on the progress here. In talking about the nonwearable market guys, is there a channel or ecosystem partner strategy to help penetrate the broader industrial customer base — is it through your customers to their customers? Or just any way that would be accelerated through a channel strategy?
Well, we're working with various partners, including a distribution partner and some reps. So we are expanding. However, as you know, those markets do take a little bit longer than the consumer market. So we are seeing a very strong, as I said, doubling year-over-year demand, but it will take a little time. But yes, we're very optimistic that those markets will grow and will be a significant part of our future revenue.
Yes. And I would say that the cool thing if I look at medical, industrial, smart home is the huge diversity of use cases that appear in our funnel, whether it's opportunities we're engaging with or wins that we have, but it's everything from Holter monitors to fetal heart rate monitoring to EEG, brain monitors, ECG patches, seizure detection devices. On the industrial side, you've got radiation dosimeters, implantable animal monitors, seabed sensing. So a huge variety of stuff. And then the other thing I'll mention from a partner side is a lot of what these customers sell are modules, right? So it's devices where we sell a chip to the module manufacturer and they go off and sell — aggregate a bunch of other customers. And in that way, we reach a broader customer base. So bottom line is we're very pleased with the diversity of customers that we're seeing, and we're optimistic about the future there.
It's growing fast outside of wearables. And I think you're going to see edge AI in a world of devices all around your personal life. So we're very confident that it's going to expand faster.
Sounds exciting. And then my second question, obviously, a great job on the fundraising here. Now that you have the stronger balance sheet, I'm wondering your thoughts and strategy on inorganic activity and if there are kind of product holes, software or hardware, that could expand? Or what the thoughts there are as you go forward, obviously strong organic growth. So wondering what the thinking is there?
We cannot talk in detail about our strategy on organic or inorganic growth. But definitely, we're using the capital to expand our portfolio of products, and it could be beyond what we talked about today around Atomiq. But please stay tuned. We're very excited. Having these kinds of funds to expand our portfolio is really exciting to us.
Your next question comes from the line of Tim Arcuri with UBS.
Jeff, I want to ask about OpEx. So it's ballooning a bit in September. The full year implies it's going to go down a touch in December. But the $7 million to $10 million worth of IP purchases, it's not something that you called out in the past. So how much of this is sort of recurring as we head into '27? Basically, what's the right baseline off of that $23 million in December and will these IP purchases continue into the first half of next year?
Well, we talked about IP purchases at the very beginning of the year. And in terms of that $85 million of OpEx spending for 2026, we said that the IP piece of it would be about $7 million to $10 million in the OpEx line. I think we're just reiterating the fact that we still are on track to spend about that much for the year, both the $85 million as well as the $7 million to $10 million for IP. That IP is directly linked to the 110 and 340 development. It's a variable cost associated with building new products. So when we think about the future, we will continue to spend money where we need to license IP to develop products beyond 340 and 110.
Okay. So it's going to recur into next year. That's the answer to the question... Then how about this? So of the year-over-year growth in revenue, so of the, let's say, $18 million September '26 versus September '25, how much of that is units versus price?
I can't give you an exact percentage, but clearly, to get that type of growth, it's unit based. I mean the demand for end customers on a unit basis is exceeding all of our expectations. And so units are the primary driver to our revenue growth.
There are no further questions at this time. This concludes today's call. Thank you for attending. You may now disconnect. Thank you.