管理層發言
Good morning. Welcome to the Amprius Technologies Second Quarter 2026 Earnings Conference Call. Joining us for today's presentation are the company's CEO, Tom Stepien; and CFO, Ricardo Rodriguez. Please note that this presentation contains forward-looking statements, including, but not limited to, statements regarding the company's financial and business performance, business strategy, future product development or commercialization, new customer adoption, and new applications, the company's growth and the growth of the markets in which it operates and the timing and ability of Amprius to expand its manufacturing capacity, scale its business and achieve a sustainable cost structure. These statements involve known and unknown risks, uncertainties and other important factors that may cause Amprius' results, performance or achievements to be materially different from any future results, performance or achievements expressed or implied in such forward-looking statements. For a more complete discussion of these risks and uncertainties, please refer to Amprius' filings with the Securities and Exchange Commission. This presentation includes a non-GAAP financial measure, which is adjusted EBITDA. This non-GAAP financial measure does not replace the presentation of Amprius' GAAP financial results and should only be used as a supplement to, not a substitute for Amprius' financial results presented in accordance with GAAP and may not be comparable to calculations of similarly titled measures by other companies. A reconciliation of adjusted EBITDA to net loss, the most directly comparable GAAP financial measure, is included in our press release, a copy of which is filed with the SEC and posted on our website. Finally, I would like to remind everyone that this conference call is being webcast. A recording will be made available for replay on the company's Investor Relations website at ir.amprius.com. In addition to the webcast, the company has also posted a press release that accompanies these results, which can also be found on the Amprius Investor Relations website. Before turning the call over to management, I want to highlight a few near-term investor relations events. Amprius will be attending the Canaccord Conference and partaking in the UBS Energy Transition Call Series next week. The team will also be attending the H.C. Wainwright Conference and the Evercore ADAS, AV, and AI Forum in September. We hope to connect with many of you at these upcoming events. I'll now turn the call over to Amprius Technologies CEO, Tom Stepien, for his comments. Sir, please proceed.
Welcome, everyone, and thank you for joining us this morning. I'm pleased to report that Amprius continues to experience robust demand for energy-dense silicon-anode lithium-ion batteries. In the second quarter of 2026, we achieved another record revenue, as we show on Slide 3. We believe the characteristics of our cells make them a particularly strong fit for one of the fastest-growing markets in the world: battery-powered unmanned aerial vehicles, UAVs, also known as drones. Our second-generation SiCore silicon-anode battery continues to gain broad adoption for drones and other applications. Given the strong quarter and promising new business, we have the confidence to increase our 2026 revenue forecast for the second consecutive quarter. Anyone who reads the news understands that low-cost drones are playing an asymmetric role in military conflicts around the world, changing the nature of modern warfare. As Barron's magazine recently reported, inexpensive drones are "upending the defense sector." We believe that this has been made possible in no small measure by the availability of high-performance batteries like those produced by Amprius. As you are likely aware, the Trump administration's proposed fiscal 2027 defense budget calls for more than $50 billion in outlays for the Defense Autonomous Warfare Group, an arm of the Department of Defense focused on drones and related hardware. The specific ask is for a 24,000% year-over-year increase. While the U.S. Congress has yet to pass the 2027 spending authorization and the actual budget might be smaller than the original request, it is nonetheless clear that the Pentagon will be making a major and growing outlay for autonomous capabilities for years to come. Our current contract with the Defense Innovation Unit gives us a front-row seat in this arena. This is good news for Amprius, our customers and partners, and for our shareholders. While we expect that the budget details will be sorted out in the months ahead, there are other positive signs from the defense sector for Amprius. For instance, the Department of Defense's drone dominance program has invited 19 drone manufacturers to a competitive demonstration event at Fort Carson, Colorado later this month. Half of the participants are using Amprius cells, and we have had at least initial conversations with the rest. The Department of Defense has said that at the completion of the Colorado event, they will place orders for 60,000 drones from the top performers. This is one more indication that the opportunity in military UAVs is in its early stages. You may recall that our contract with the Defense Innovation Unit to develop and scale National Defense Authorization Act, NDAA-compliant silicon-anode battery cells was increased for a third time in the March quarter and now totals $18.1 million. This funding supports expansion of a pilot line at our headquarters in Fremont, California. I'm happy to report that we have now received nearly half of the equipment required for the pilot line, which is undergoing installation. The remaining equipment is scheduled to arrive later this month and in September with production expected to begin in December 2026. Our opportunity in military drones goes beyond U.S. borders. I'm happy to report that we've received a $24 million order from a new European customer, a drone manufacturer that uses our SA124 SiCore cylindrical cells. We started to deliver our batteries to this customer in Q2 and will continue to do so for the next three quarters. We're excited about this opportunity and expect to have more to talk about on this topic in the months ahead. I also want to report some recent news from our customer Redwire, a leading aerospace and defense technology company providing space infrastructure, autonomous systems, and mission-critical solutions for commercial, civil and national security customers worldwide. Redwire first purchased our high energy density batteries in 2024. We're happy to see that their demand is growing. Redwire recently announced more than $40 million in purchase orders from the U.S. Marine Corps for the Stalker Block 30, a Group 2 drone designed for long-range reconnaissance. Stalker excels in missions where long endurance is critical, and that's exactly what our SiCore cells enable. We look forward to Redwire's continued success. While we are excited about the rapid adoption of drones in the defense industry, we also continue to see substantial opportunities for commercial drones. In May, we announced an agreement to provide high-density silicon-anode cells to Matternet, the world's only FAA-Type Certified drone delivery platform. Amprius' silicon-anode cells deliver up to twice the energy density of conventional graphite-based batteries, a critical advantage in aviation where low battery weight and high energy density directly improve aircraft range, payload and economics. Drones, both military and commercial, are a big part of the Amprius story, but they are not our only strategically important end market. Earlier this year, we announced a $21 million order from a premier electric mobility customer in China to power a suite of light electric vehicles, including scooters, three-wheelers and motorcycles. Today, I'm excited to tell you about a new e-mobility customer win. We have signed a three-year contract with Stark Future, a Barcelona-based premium electric motorcycle manufacturer. Stark's bikes are impressive, with cutting-edge technology, world-class design and great engineering. We showcased one of their bikes at our CES booth in January 2026. We expect revenues from our relationship with Stark to be at least $100 million through 2029, with shipments expected to start early next year. Let's turn to Slide 4 and discuss our go-to-market strategy. A little less than half of our sales ship directly to end-use customers—companies that sell drones for various applications, small electric vehicles, and companies in the satellite value stream. We have more than 500 direct customers, and this figure grows every quarter. The other portion of our purchase orders come from PAC partners—companies who buy our cells and package them together with appropriate electronics and sometimes a battery management system. These modules and packs are then sold to end-use customers. Our PAC partner program allows us to expand our reach and simplifies customer relationships for many end market applications. Today, we list nine PAC partners on our website and more will be added over time. This program is a light lift for our sales team and allows us to scale without adding direct sales headcount. We expect this flywheel effect to allow us to continue strong growth through this rapidly expanding channel. Let's turn to Slide 5, and let me provide an update on our capital-efficient contract manufacturing strategy. Our California pilot line gives us the ability to win new customers, allowing us to quickly deliver new cell chemistries to customers. We often do this side-by-side with our customers in joint development programs. We also use our Fremont facility to deliver small volumes of cells. We leverage our worldwide contract manufacturing partners to produce sales at volume. We have four manufacturing partners in China. Over the last several years, they have provided us with excellent quality and reliable delivery. Our China contract manufacturers have helped Amprius tremendously, and they will continue to be important partners in our future. We have added several partners in South Korea. In May 2025, we announced our first partner, Libest, located in Daejeon, about a two-hour drive from downtown Seoul. Libest has been delivering Amprius commercial cells since September 2025. We recently added JR Energy and Top Material as South Korean partners. Together, these three contract manufacturers give us the ability to produce batteries that are compliant with NDAA rules. I met with all three partners in Korea ten days ago and can confirm that our relationships are solid, our incentives are aligned, and we are expanding well together. I also want to underscore that we believe we are well on track to reach full NDAA compliance with domestically produced cells in 2027. South Korea gives us NDAA-compliant supply today, while Nanotech Energy, the U.S. contract manufacturer we announced earlier this year, provides additional U.S. capacity. We expect to talk about additional U.S. partners in the months ahead. Our partner-focused approach to manufacturing avoids substantial capital expenditures while keeping our management team focused on what matters most: extending our technical innovations and delivering these innovations to customers rapidly. Leveraging manufacturing partners rather than investing heavily in new facilities allows Amprius to scale quickly and efficiently while maintaining strategic flexibility. I want to provide a brief update on our senior management team. Last month, Ronnie Tao, a five-year Amprius veteran who until recently served as our VP of Sales, moved into a new role as Chief Business Officer. Ronnie will focus on expanding our reach into new markets, initially targeting robotics, where we see substantial opportunity for our high-energy density offerings. Ronnie's energy is infectious and his recall of technical details is remarkable—ideal qualities to drive growth into new segments. Three weeks ago, Anne Torricelli joined Amprius as our new VP of Sales. Anne has nearly two decades of experience working in energy technology sales and business development roles. She was most recently Managing Director of Energy Storage Solutions for Gotion, a top-five worldwide lithium-ion battery manufacturer. Her multicultural background, excellent communication skills, and savvy organizational traits are a model profile to lead sales for a fast-growing international company like Amprius. We're thrilled to welcome Anne to the team. A few additional thoughts before I pass the microphone to Ricardo Rodriguez, our CFO. I continue to see wide and varied growth opportunities for Amprius in multiple markets and several geographies. As I noted earlier, there are sizable opportunities for us in drones, not only in defense, but also for commercial delivery, public safety, security, and a growing number of other applications. The massive commitment to drones from the Department of Defense is a positive sign, but it's not the only one, and we expect drones to expand across many parts of the economy in years ahead. As highlighted by our new relationship with Stark in Spain, there is also a fast-growing opportunity for battery-powered mobility, including robotics. We are focusing some of our key executives on this emerging market for personal and commercial robots, including delivery bots, humanoid, and industrial mobile robots. It is early, and there is no meaningful robotics revenue in our numbers today. But our ability to offer power and energy-balanced cells plays well in the unstructured environments in which these machines operate. We expect to have more to say about this segment in the months ahead. Another opportunity that exists for us is in satellites and space, where our high energy density cells directly improve launch economics. Satellite launch providers charge customers by weight, making our ability to deliver the same energy at roughly half the weight extremely valuable. Finally, we believe there is tremendous potential for our batteries in eVTOL—electric vertical takeoff and landing aircraft—for autonomous point-to-point regional transport for both passengers and cargo. It's early in development of eVTOL vehicles, but they are coming sooner than many think. Let me now turn the call over to Ricardo to review our Q2 results in detail.
Thank you, Tom, and good morning, everyone. I'm happy to start on Slide 6. In the second quarter, we delivered $34 million of revenue, up 19% from the first quarter and 2.3x year-over-year. This was our sixth consecutive quarter of sequential growth, and it puts our annual revenue run rate at $136 million. We expect this to continue growing, so more on this later. For the first half of the year, revenue was $62.6 million, up 137% year-over-year. SiCore accounted for 98% of our revenue in Q2. Regionally, EMEA drove 68% of our revenue, with the rest of our revenue coming from the U.S. and Asia. Cost of goods sold was $24.8 million, up 9% against 19% revenue growth, which enabled gross profit of $9.3 million and a gross margin of 27%, right in line with our expectations and improving from 20% in the first quarter and 9% in the same quarter last year. For the first half, gross margin was 24%, improving from negative 4% in the first half of 2025. Total operating expense was $13.6 million in the quarter, up $1.2 million sequentially and $5.4 million year-over-year. We continue investing in our go-to-market and R&D efforts as these continue to pay off as our team wins in the market. Our operating loss in Q2 was $4.3 million compared to $6.7 million in the first quarter and $6.8 million in the same quarter of last year. Other income was $1.1 million, consisting of $472,000 of interest income and $700,000 of government grant income tied to our work with the Defense Innovation Unit. Our GAAP net loss attributable to common shareholders for the second quarter was $5.1 million or negative $0.04 per share based on 143.5 million weighted average shares outstanding. That is 20% narrower than the same quarter last year. For the first half, our net loss was $10.1 million compared to $15.7 million in the first half of 2025. Our GAAP net loss includes a one-time noncash $1.9 million adjustment reflecting the change in fair value of the public warrants during our exchange for stock on May 6 of this year. Excluding this $1.9 million gives us non-GAAP adjusted net loss of $3.2 million or negative $0.02 per share for Q2 and $8.2 million or $0.06 per share for the first half of 2026. Adjusted EBITDA in the second quarter was negative $1.0 million or negative 3% margin compared to negative $1.8 million in the first quarter and negative $2.1 million in the same quarter last year. As a reminder, we define adjusted EBITDA as net income or loss before interest, taxes, depreciation, amortization, stock-based compensation and other items that we do not believe are indicative of our core operating performance. In the second quarter, these adjustments were limited to the $1.9 million related to the warrant exchange, $2.5 million of stock-based compensation, $800,000 of depreciation and amortization and $1.1 million of interest and other income. For the first half of the year, adjusted EBITDA was negative $2.8 million against negative $7.3 million in the first half of last year. On a trailing 12-month basis, adjusted EBITDA is negative $800,000 with a negative 1% margin. We are within a rounding error of breakeven on a full-year basis if we look at the last 12 months. Now turning over to cash flow and the balance sheet. We ended the second quarter with $74.5 million of cash and no debt, an increase of $12.2 million during the quarter. Our operations only used $2.9 million of cash in the quarter. Accounts receivable grew by $5.4 million and inventory grew by $3.3 million, partially offset by lower prepaid inventory and higher payables. Before those working capital movements, our operations generated cash for the first time. Accounts receivable ended at $40.7 million and inventory at $11.5 million. Both are deliberate. Receivables reflect a fast-growing shipment profile weighted towards the second half of the quarter and an inventory position to support the ramp of the third quarter. Capital expenditures were $1.8 million, all at our Fremont facility, supporting the electrode coating build-out and primarily funded by the Defense Innovation Unit. First half CapEx was $2.8 million against the less than $10 million that we framed out for the year in March, and we are tracking well inside that. Financing activities provided $16.8 million, consisting of $12.3 million from warrant exercises and $4.5 million from option exercises. We currently do not have an at-the-market offering program. Every dollar of equity capital we took in this quarter came from holders choosing to exercise into the business as we continue to focus on minimizing dilution. Working capital at quarter-end was $113.2 million, compared to $59.8 million for the second quarter of last year, and total stockholders' equity was $125.6 million. Before I turn the call back to Tom, I want to frame our outlook for the rest of the year using Slide 7 as the backdrop. In March, we set an initial 2026 outlook of more than $125 million of revenue, over 25% gross margin and our first full year of positive adjusted EBITDA. In May, after a first quarter of $28.5 million of revenue, we increased guidance to more than $130 million for the year. Today, we are increasing our revenue forecast again. With what we know today, we expect full year revenue of at least $140 million and gross margins of at least 28%. We see upside to gross margins in the second half of the year as we focus on ensuring that the fixed cost of our contract manufacturing partners does not increase with higher volumes. We are reiterating adjusted EBITDA of more than $4 million, a net loss of $10 million or less and a loss of $0.08 or less per diluted share, assuming 143.5 million weighted average diluted shares. These updated GAAP profit guidance estimates consider the $1.9 million adjustment for the fair value of the warrants in Q2 of this year. Looking further ahead, nothing about the plan that we laid out in March has changed, except for how much of it is now visible in the numbers and in reality. As we close out the decade, we are still targeting more than $600 million of contracted capacity, gross margins above 30% and adjusted EBITDA margins of at least 20%. The resourceful culture and low fixed cost structure that brought us within rounding error of breakeven over the last 12 months are the same ones that will get us there. With that, I'm happy to turn the call back to Tom for his closing remarks. Thank you very much for your attention and continued support.
Thanks, Ricardo. We remain excited about the opportunities ahead and look forward to meeting many of you at upcoming investor events. Thank you for your continued interest and support of Amprius. And with that, let me turn it over to the operator for questions.
分析師問答
The first question comes from Colin Rusch with Oppenheimer.
Congratulations on the progress here. As we see some of the incremental regionalization and the NDAA compliance mandates starting to flow through, I want to get a sense of how much dexterity the technology platform has in terms of using alternate inputs on the anode side as well as on the electrolyte side, whether you're moving towards a semi-solid-state or solid-state electrolyte at some point, and your ability to actually integrate some of those material advances that we're seeing in the ecosystem.
Yes, Colin. We have five chemistry platforms: a power-based one, an energy-based one, a balanced one, among others. We have very good dexterity. We have long-serving suppliers over the last several years. As part of NDAA work, we've qualified two sets of 11 new suppliers for the anode, cathode, separator and seven other binders and related components that go into our batteries. We have a primary set of new suppliers and a secondary set as well. We are using those suppliers in the cells that are being made in South Korea and the U.S., so it's working well. It is tricky because we have to qualify them and get these suppliers under contract, so it keeps us busy on the supply chain and operational side of things as well as the technical side. The pilot line here in Fremont helps because we get quick turns and can rapidly validate some of those new components. We're pretty nimble in that area. We'd always like to go faster, of course, but we're happy with where we are, and we're on track, as we said in the call, to be fully NDAA-compliant with U.S. production in 2027.
Excellent. And then shifting to the customer base: given that level of range from the platform, over the last 1.5 years you've done a great job of actually getting a lot of these customers organized more methodically in terms of their purchasing patterns. But I want to get a sense of, as you move forward leveraging the technology roadmap, some of the range of possibilities and scale that you guys can leverage into both driving incremental sales as well as operating margin—how should we think about that playing out over the next 12 to 24 months?
The PAC program certainly helps, as we tried to say in the call, giving us some leverage in the flywheel effect. That helps. Anne joining us is a very organized, methodical person who has run large sales groups, and that will help. The breadth of our offerings allows us, without a lot of complexity, to serve new segments. Robotics and eVTOL are early, but we have a couple of joint development programs underway. The tweaks we're making side-by-side with future customers are relatively small and can be done rapidly to provide the energy and power characteristics they need. That will help us go deeper as well as wider on the customer side.
The margins really depend on both the regional mix and the product mix. We see our sales being most accretive for us, and as we launch NDAA-compliant pouch cells, we think that will be accretive and help us get to our margin targets. When we reported Q3 last year, we said the margins would be lumpy, and that happened: margins were 20% on a GAAP basis in Q1 and 22% if you strip out one-time costs from the Colorado facility. In this quarter, the revenue mix was favorable with significant sales in Europe and with pouch cells holding their share. As we look at markets like robotics and eVTOL, and a team now looking at potential data center applications, we believe these will be pursued in an accretive way.
The next question comes from Mark Shooter with William Blair.
Tom, we saw the L3Harris order and the Nanotech contract manufacturing as a creative way to find and leverage some small U.S. manufacturing capacity that's available for domestic supply-chain sensitive customers. As stricter domestic requirements from the Department of Defense kick in in '27 and '28, do you see this as an inflection point for domestic drone manufacturing or even for cell manufacturing? Will you have to ramp up more U.S. manufacturing? If so, what avenues do you see to do so?
We have pretty good visibility through the next several quarters, and on a planning basis we look through the end of 2027 and 2028 with increasing certainty. We know when these stricter requirements kick in. We're pleased with where we are in Korea with the three partners. We need to go faster and deeper. Nanotech has delivered cells; we need to flow more through them. We're actively working with other U.S. manufacturers; we are not quite ready to announce additional partners yet, but stay tuned. We always want to move faster, and we are pleased with demand and our technical leadership, but we need to earn it every day by delivering to requirements like L3's.
Great. I appreciate the color. Ricardo, can you give us a little more color on what you're seeing that gives you the confidence to raise both revenue and margin guidance? What are the key drivers you are seeing?
The revenue increase was straightforward given the demand profile. As the team develops tighter relationships with some of the pack houses and several OEMs, we are seeing more consistent order flow to our customers. That gives us visibility into demand for the second half and into next year. There's upside to the guidance, but updating it to the new level makes sense given the supply picture and how the supply chain needs to evolve to deliver product in Q4 and into next year. The margin raise was also straightforward. If you look at our margins for the first half and strip out roughly $0.5 million of expenses from Colorado in Q1 and roughly $1.8 million in tariffs that have effectively been refunded, our margins would have been 27% for the first half. Raising to at least 28% in the second half is supported by a better understanding of the fixed-cost elements inside our contract manufacturing partners. It would be unfair to let those fixed costs scale up at higher volumes without addressing them. To hit this guidance, we do need a favorable mix—at least 60% of revenue from Europe and a majority from accretive sales—and we see that playing out for the rest of the year.
The next question comes from Eric Stine with Craig-Hallum.
Can we go back to NDAA compliance? It sounds like you've made progress on the 11 needed components with primary and secondary suppliers. You mentioned getting those suppliers under contract. Last quarter you indicated progress—would love an update on that.
We make progress on a weekly basis with the suppliers. I met with several when I was in Korea a couple of weeks ago. There have been many discussions since getting back. We have a team working on it and some outside resources helping accelerate the process. We'll get them all buttoned up under contract this quarter. Legal reviews tend to take time, but we're comfortable with the cost and timing. These new suppliers are large international companies with significant revenue, so we are comfortable with their ability to deliver. The complexity is primarily in operational planning and aligning schedules, since they're shipping products to different manufacturers in different countries. It's not technically hard, but it is somewhat complex on the planning side.
Okay. And one more: on the commercial side, DoorDash recently got Part 135 FAA certification for drone delivery. Going forward, do you envision certification as company-by-company, or could the FAA open broader pathways? There's a sizable commercial drone opportunity here.
We have a start with Matternet, as we talked about on the last call, and that is a foothold as we begin to expand. I don't know that the FAA certifies batteries; they tend to certify vehicles on a vehicle-by-vehicle basis. We're not FAA experts, but this is part of why we made changes in our sales team—Ronnie and Anne—to focus on these opportunities. The new delivery drones won't necessarily drive the next quarter's revenue immediately, but planting seeds and understanding their requirements is part of Ronnie's remit so we can turn those seeds into future revenue. If delivery vehicles can use our batteries, they could go further or carry greater payloads—many payload limits are a couple of kilos and our technology could potentially double that. The promise is there; we have work to do to earn the trust of companies like DoorDash and others.
The next question comes from Austin Bohlig with Needham & Company.
Congrats on the great results. Tom, can you dive into the key geographies and outlook driving the upside? Is the momentum more domestic in the U.S. or international?
International has been strong—Europe especially. Last quarter Europe was 75% of revenue; it was a bit lower this quarter but still a solid base. We're happy with Redwire and other customers coming on. We see some shifts from fiscal government spending in 2026 starting to flow; some customers received funds that started earlier. We like the balance Stark brings in e-mobility as a complement to drones. Over time, we expect a better balance between Europe and the U.S. and Asia, with growth across e-mobility and drones, including both defense and commercial delivery and public safety.
Super. One follow-up: you mentioned half of the drone dominance customers are using SiCore. Could you walk through the content per drone opportunity? I know there were slides in the deck, but any clarity on this program's potential would be helpful.
We included helpful detail on Slides 10 and 11 in the deck. Different drone groups use different amounts of battery content: Group 1 uses less, Group 3 tends to use more, and Groups 4 and 5 are often fuel-based. Those slides provide typical battery sizes and potential cell content by class. The nice thing with our cells is they tend to be sticky—once customers adopt our batteries, they often stay with them. They value increased flight time, and in shootouts higher energy density often scores very well. That's how we view DDP and similar opportunities.
Refer to Slides 10 and 11 in the deck for additional detail on battery sizes and content by drone class.
The next question comes from Ryan Pfingst with B. Riley.
Tom, you talked about robotics earlier. What is Amprius doing today to position itself with customers as that market scales commercially?
Step one is getting smarter about the market. We attended Automate in Chicago recently—our first robotics show. We learned that our cells perform well for robots operating in unstructured environments—dog-like robots and certain humanoids that operate outside of tightly controlled factories. For robots that lift or have current spikes, specific battery power characteristics are needed; for robots that wander without certainty of when they'll charge, longer run time and high energy density are prized. That insight allows us to focus our efforts. Executive changes—Ronnie leading business development and Anne in sales—are intended to help us plant seeds and convert them into revenue as the robotics market matures.
Appreciate that. Ricardo, you mentioned a team looking at data centers. Can you talk more about that and where Amprius could fit in?
It's an initial look, but the power requirements of modern CPUs keep increasing. Operators use supercapacitors to smooth high-power draws, but you might need a high-power cell immediately behind those supercapacitors to help flatten the load. We see an opportunity for a high-power cylindrical cell located close to the rack, providing 1 to 60 seconds of runtime, recharging, and then being ready again for the next event. We're having initial discussions with integrators of racks and compute infrastructure, and we'll report back as we get customer traction.
The next question comes from Tim Moore with Clear Street.
Nice execution in the quarter. You have the potential for high operating leverage on SG&A over the next few years. The gross margin guidance hike is impressive. Looking at your 2030 target of above 30% gross margin, how should we think about possibly pulling that target forward? Could you get to 30% plus in 2029?
We will try to do it sooner, and the elements are there, but we want to be cautious about what we sign up for. Over the next 12 to 24 months, we are managing the full NDAA compliance, which comes with a different cost structure that we need to be paid for. That's why we put the 30% plus gross margin target out to 2030. We could pull it ahead if the mix is favorable. The main driver is revenue mix—if we sell more customized pouch cells, that helps; if more cylindrical content comes from non-accretive markets, it could be less favorable. There will be lumpiness in quarters as mix shifts or as North America revenue increases before full compensation for NDAA-related expenses flows through, so it will be a journey.
Second question: thinking about U.S. pouch manufacturing—are you close to purchasing an existing facility in the U.S. to retool for pouch cell manufacturing? That could help U.S. manufacturing. Curious about potential timing.
You're reading our mind. We're looking for a partner to do that; there's capacity from firms that leaned into EVs aggressively over the past few years. Stay tuned.
To clarify, it likely will not be Amprius purchasing a facility. Our model is to work with partners. Whatever we do, whether in the U.S. or elsewhere, will likely be done with partners. Ricardo's point is correct—there are partner options and capacity in the U.S.—stay tuned.
Also, note that not all the CapEx is the same. The equipment inside the facilities can be a smaller share of the total investment than building the facility itself. We learned from Colorado that equipment can be about one-third of what building the facility costs. If you can structure the business so equipment payback works and ideally not have to fund the building, you can get payback in two to three years if you fill the capacity. We're getting signals that demand will be there as we talk to the Department of Defense; we just need to put the pieces together to deliver it.
The next question comes from Chip Moore with ROTH.
I wanted to go back to the drone dominance program. Half of those finalists are using SiCore—great testament to performance. Can you expand on the competitive environment and the alternatives out there?
It's very competitive. We tend to win because of energy density. These shootouts or gauntlets are scored on metrics like duration—how far and how long drones can fly—so higher energy density is generally better. There's a mix of cylindrical and pouch cells. Fully optimized drones tend to use pouch for greater energy density, but pouch packs are a bit harder to integrate compared to cylindrical cells. Our sales team focuses on those trade-offs. We have a malleable platform with five different chemistries, so we can get something in front of customers to earn their trust and win business. We're in about half of the 19 participants for the upcoming event, but we have more work to do.
And on go-to-market, can you discuss visibility differences between PAC partners and OEMs and how you see that mix evolving over the next two to three years?
PAC partners are an extension of our team. Some drone companies rely on PAC partners for recommendations, and we want to be at the top of their lists. We are building that ecosystem—nine partners are listed on our website—and we think of it as a gold/silver/bronze structure as we go deeper. Visibility is improving: some companies can't give multi-quarter purchase orders because they lack visibility, but as customers like Redwire get under contract with their customers, our visibility increases. We have very good visibility for Q3 and Q4; it gets grainier farther out into 2027. It's been improving but we have more work to do.
The next question comes from Derek Soderberg with Cantor Fitzgerald.
What's your current backlog? I know backlog isn't always the best metric for you since you ship a lot in-quarter. Also, can you explain margin differences between defense versus UAV cylindrical and the light electric vehicle opportunity—what's the range on margins there?
On backlog, there's not much to add—historically the word backlog has been misused. We do have visibility into open purchase orders and feel good about where we stand as we work through Q3. We see similar dynamics to recent quarters: POs arrive and we work to deliver as much as possible within the quarter. For margins, the range is roughly low 20s to high 30s percent depending on form factor and region. On the lower end are more commoditized cylindrical cells that are easily swappable; on the higher end are highly customized pouch cells. Region also matters—some China sales could be lower margin, even single digits in certain instances, and we're careful about playing in those segments. The same range applies across defense and non-defense applications; there isn't significant pricing discrimination between end markets, so form factor and region drive the spread.
That's helpful. How much of current revenue is cylindrical SiCore? Regarding the $24 million European cylindrical order, how much of that will you ship in 2026 versus 2027?
The $24 million order should be done over the next two and a half quarters. We haven't broken out the exact mix between pouch and cylindrical publicly because we view that as competitive intelligence.
The next question comes from Ted Jackson with Northland Securities.
You have China capacity and South Korean capacity and the beginnings of U.S. capacity. Can you refresh us on what the production capacity is in each region, and maybe refine it between cylindrical and pouch? What do you have now and where can it go?
We total our contracted capacity in the deck to greater than 2 gigawatt-hours per year. In earlier calls we broke that out more granularly; on average that equates to roughly 12 to 15 million cells per quarter on a sellable unit basis. We do not break out capacity by contract manufacturer publicly. We named the three CMs in Korea and noted Nanotech in the U.S., and we are working to identify U.S. pouch manufacturers. We aim to have supply slightly ahead of demand so we don't leave revenue on the table. China is relatively balanced between cylindrical and pouch; that balance is emerging in Korea. In the U.S., we are currently unbalanced—identified for cylindrical but working to add pouch. The aim is always to have extra capacity to meet demand and win through speed and nimbleness.
There is nuance: our contract manufacturing partners do not want idle capacity, and in one way or another, we would be paying for it. Internally we manage supply like Tetris: demand comes in a mix of SKUs and regions, and within each period we work to ship as much as possible given our supply sources and SKUs. We've increased flexibility in South Korea and are looking at creative ways to produce cylindrical cells in NDAA-compliant markets to increase coverage and flexibility. That's how we match supply to demand in the near term.
What's the typical lead time for planning capacity with manufacturers? How do you coordinate and understand available capacity from your contract manufacturers?
Lead time depends. For full NDAA compliance and setting up U.S. capacity, lead times can be a year or two. We work well in advance of need by understanding demand and then farming out capacity and setting up commercial relationships. If we're talking about flexibility within region, we can be much faster and make switches within a quarter to one and a half quarters. That's how we've managed to be flexible over the last three to four quarters.
The next question comes from Amit Dayal with H.C. Wainwright.
Tom, regarding the 2027 spending authorization, it looks like it could get done by December. Should we expect contributions from this starting in 1Q 2027, or is there a lag between approval and when you begin to see orders?
There likely will be a lag. We've seen that historically—customers receiving defense funds often occurs later in a fiscal year, and there's typically a six- to nine-month lag from budget availability to flow down to companies like Amprius. If the budget is resolved on the timeline you mentioned, some of that will trickle in mid-2027 with other orders following after that. So expect at least a quarter's lag. The large ask to the Department of Defense bodes well, though not all of it will come through, and we want to be ready to deliver quickly to those customers.
On receivables, you've grown AR to over $40 million. Are you comfortable with collectability and any working capital pressure as you grow rapidly?
Yes, we're comfortable. Going into the year, we planned for roughly $20 million to $25 million of working capital to enable a doubling of revenues. The receivables balance largely reflects the snapshot at quarter close; if you looked a month after quarter close, you'd see many receivables collected. The AR aging is driven largely by large, long-standing customers. We also built inventory intentionally to enable the second-half ramp. From a working capital perspective, we're comfortable with the position.
At this time, I would like to turn the call back to management for closing comments.
To all our customers, shareholders, employees and partners, thank you for your continued support. We are at an opportune moment. Today, several markets require advanced batteries and Amprius cells lead the industry on multiple dimensions. That intersection is a powerful one, and we expect to continue to drive technical innovation, execute with discipline, and deliver meaningful results. Thank you for your time and attention this morning.
Thank you. This concludes today's teleconference. You may disconnect your lines at this time. Thank you for your participation, and have a great day.