AXT, Inc. Aktie

Good afternoon, everyone, and welcome to AXT's First Quarter 2026 Earnings Conference Call. Leading the call today is Dr. Morris Young, Chief Executive Officer; and Gary Fischer, Chief Financial Officer. In addition, Tim Bettles, VP of Business Development, will be participating in the Q&A portion of the call. My name is Tracy, and I will be your coordinator today.

[Operator Instructions] I would now like to turn the call over to Leslie Green, Investor Relations for AXT. Leslie, go ahead.

Thank you, Tracy, and good afternoon, everyone. Before we begin, I would like to remind you that during the course of this conference call, including comments made in response to your questions, we will provide projections or make other forward-looking statements regarding, among other things, the future financial performance of the company, market conditions and trends, emerging applications using chips or devices fabricated on our substrates, our product mix, global economic and political conditions, including trade tariffs and import and export restrictions, ability to obtain China export permits, timing of receipt of export permits, our plan to list our subsidiary, Tongmei in China, our ability to increase orders in succeeding quarters to control costs and expenses, to improve manufacturing yields and efficiencies or to utilize our manufacturing capacity.

We wish to caution you that such statements deal with future events are based on management's current expectations and are subject to risks and uncertainties that could cause actual events or results to differ materially.

In addition to the matters just listed, these uncertainties and risks include, but are not limited to, the financial performance of our partially owned supply chain companies and increased environmental regulations in China.

In addition to the factors just mentioned or that may be discussed in this call, we refer you to the company's periodic reports filed with the Securities and Exchange Commission. These are available online by link from our website and contain additional information on risk factors that could cause actual results to differ materially from our current expectations.

This conference call will be available on our website at axt.com through April 30, 2027. Also, I want to note that shortly following the close of market today, we issued a press release reporting financial results for the first quarter of 2026. This information is available on our website at axt.com.

I would now like to turn the call over to Gary Fischer for a review of our first quarter results. Gary?

Thank you, Leslie, and good afternoon to everyone. Revenue for the first quarter of 2026 was $26.9 million compared with $23.0 million in the fourth quarter of 2025 and $19.4 million in the first quarter of 2025 last year.

To break down our Q1 '26 revenue for you by product category, Indium Phosphide was $13.6 million, primarily from data center applications. Gallium Arsenide was $5.4 million, Germanium substrates were $200,000. Finally, revenue from our consolidated raw material joint venture companies in Q1 was $7.6 million.

In the first quarter of 2026, revenue from Asia Pacific was 78%, Europe was 21% and North America was 1%. The top 5 customers generated approximately 32% of total revenue and no customers were over the 10% level. Gross margin showed a substantial improvement in the first quarter.

Non-GAAP gross margin was 29.9% compared with 21.5% gross margin in Q4 of 2025 and a negative 6.1% gross margin in Q1 of 2025 last year. For those who prefer to track results on a GAAP basis, gross margin in the first quarter was 29.6% compared with 20.9% in Q4 and a negative 6.4% in Q1 of 2025.

Moving to operating expenses. Total non-GAAP operating expense in Q1 was $8.6 million compared with $7.5 million in Q4 and $8.5 million in Q1 of 2025. On a GAAP basis, total operating expenses in Q1 was $9.6 million compared with $8.7 million in Q4 of 2025 and $9.0 million in Q1 of 2025.

Our non-GAAP operating loss for the first quarter of 2026 was $550,000 compared to a non-GAAP operating loss in Q4 of 2025 of $2.6 million and a non-GAAP operating loss of $9.6 million in Q1 of 2025.

For reference, our GAAP operating line for the first quarter of 2026 was a net loss of $1.6 million compared with an operating loss of $3.8 million in Q4 of 2025 and an operating loss of $10.3 million in Q1 of 2025.

Nonoperating other income and expense and other items below the operating line for the first quarter of 2026 was a net loss of $35,000. The details can be seen in the P&L included in our press release today.

In Q1 of 2026, we made substantial progress towards profitability. We had a non-GAAP net loss of $585,000 or $0.01 per share compared with a non-GAAP net loss of $2.3 million or $0.05 per share in the fourth quarter and non-GAAP net loss in Q1 of 2025 of $8.2 million or $0.19 per share.

On a GAAP basis, the net loss in Q1 was $1.6 million or $0.03 per share compared to a net loss of $3.6 million or $0.08 a share in the fourth quarter and $8.8 million or $0.20 per share last year in Q1 of 2025. The weighted average basic shares outstanding in Q1 of 2026 was 53.3 million.

Cash, cash equivalents and investments decreased by $5.1 million to $123 million as of March 31. By comparison, at December 31, it was $128.4 million. Accounts receivable increased by $5.2 million, almost exactly the same as the change in cash.

Depreciation and amortization in the first quarter was $2.4 million. Total stock comp was $1.0 million. Net inventory was up approximately $8.5 million for the first quarter to $90.2 million. This concludes the discussion of our quarterly financial results.

Turning to our plan to list our subsidiary, Tongmei in China on the STAR market in Shanghai. We remain very interested in completing the IPO, particularly in light of the rapidly evolving AI infrastructure build-out in China and China's development of its semiconductor supply chain, which is fueling increased China-based demand for Indium Phosphide substrates.

We have continued to keep our IPO application current, and Tongmei remains in process as a part of a much more selective and smaller group of prospective listings than a few years ago. Though the current geopolitical environment is dynamic, Tongmei is considered a Chinese company and continues to be regarded in China as a good IPO candidate. We will keep you informed of any updates.

With that, I'll now turn the call over to Dr. Morris Young for a review of our business and markets. Morris?

Thank you, Gary. This is an incredibly exciting time for AXT. As many of you are aware, last week, we completed a capital raise for $632.5 million in support of Tongmei's Indium Phosphide capacity expansion as well as R&D investment in new products like 6-inch Indium Phosphide and other working capital needs. With our backlog of orders and customer forecast achieving record levels, we are laser-focused on adding capacity to support customer requirements.

I'm pleased to report that we are running ahead of our plan to double our Indium Phosphide capacity this year from Q4 of 2025 levels. Our capability to scale up quickly is unique among our peers. Unlike our competitors, AXT designs and builds our own crystal growth furnaces, has our own supply of critical raw materials and has the manufacturing space in place to achieve our expansion goal this year.

As you can imagine, longer-term capacity planning is one of the most important discussions we're having today with customers and major supply chain players in our space. The message we are having for them is this.

AXT is stepping up beyond our 2026 capacity expansion, we're planning to double our Indium Phosphide capacity again in 2027, with a new facility near our current one that we will be dedicated to Indium Phosphide wafer production.

Our 2028 planning is also underway, and we expect to expand again meaningfully. This is an industry in which scale matters. The barriers to entry are high even for most skilled manufacturers. As the market continues to grow, capacity has become a critical enabler.

What we are hearing from the industry sources and echo from our customers is the expectation that the market for optical components will increase significantly in the coming years, driving a 4x to 6x increase in substrate market overall in the next three to five years, driven by both scale-out and scale-up applications.

Beyond pluggable transceivers, we are seeing a very large developing market for CPO, co-packaged optics. We are actively engaging in discussions with customers about their technical and timing requirements and believe this could represent another inflection point in our business beginning in late 2027 and beyond.

With this massive growth cycle ahead of us, we are actively working with a multitude of players from our direct customers to the end customers with whom we have not historically had direct relationships. We are there to understand their longer-term requirements and to align our growth and innovation plans accordingly. This will be a thoughtful and measured process, but we believe we are in a best position competitively to support and enable our industry in meeting its current and future needs.

Over the last few quarters, the expansion of our Indium Phosphide customer base has been gratifying. We're now supporting nearly all leading customers in the optical space. This includes Tier 1 laser manufacturers and optical transceiver module makers, both around the globe and in China.

In alignment with our customers' technical requirements and roadmaps, we're making important progress on our 6-inch Indium Phosphide product for both iron doped and sulfur doped specifications. A significant part of our capacity expansion will be focused on 6-inch crystal growth technology to support the planned roadmap of 6-inch capability by our customers. We're excited to be able to demonstrate the technological advantage of our low EPD wafers as the market moves to optical devices with higher speeds and greater sophistication for both scale-up and scale-out applications.

Now turning to Q1. Export permits in our first quarter came in slightly better than our guidance and are off to a solid start in Q2. Gary will take you through our full guidance in a few minutes, but we're expecting to achieve sequential revenue growth in Q2, driven primarily by growth in Indium Phosphide. In fact, Q2 will be expected to be our largest quarter for Indium Phosphide in AXT's history. This derives from an Indium Phosphide backlog that has now reached a new high of over $100 million.

As we mentioned last quarter, customers are giving us more visibility into their expected demand and working closely with them in this supply-constrained environment to meet their need, as we continue to expand our capacity. From a geographic demand perspective, the massive AI infrastructure build-out and planned CapEx spending by cloud services and AI platform providers in the U.S. is the primary driver for EML and silicon photonics-based optical transceivers as well as high-speed photodetectors.

We believe that today, our material is being used in multiple U.S. hyperscalers. We expect that end customers' use will continue to broaden. We're also seeing significant growth in China, as China moves to accelerate its capability throughout the AI supply chain. Our revenue related to Indium Phosphide-based laser market in China more than doubled in Q1 from the prior quarter, and we expect them to double again in Q2. This highlights China's increasing investment in AI infrastructure supply chain for the global market. This is a great opportunity for AXT as there is no permit required to ship our product within China.

Turning to Gallium Arsenide. In Q1, demand for semiconducting wafers for industry robotics and data center lasers applications all held steady from the prior quarter. We continue to see demand for semi-insulating wafers for wireless RF devices and believe that we have a strong opportunity for market share expansion. However, this is gated primarily by our ability to obtain export license, which came in light in Q1.

Finally, our raw material business continue to be a crown jewel in our growth strategy. We're pleased to report that our subsidiary, JinMei, has begun to refine high-purity Indium, which gives us direct control of a guaranteed supply of another critical material for our Indium Phosphide substrates.

We're also investing to help JinMei expand its capability so that our AXT demand grows, JinMei will continue to provide a meaningful portion of our raw material requirements. Globally, there continues to be a greater awareness of the importance of earth materials, and we are decades ahead of the curve in developing our unique integrated supply chain. We continue to invest in our portfolio as we believe it's a major competitive differentiator.

In summary, we believe AXT is entering one of the most consequential chapters in our company's history. The investment we are making today in capacity, in technology and in our unique integrated supply chain positions us to meet the extraordinary demand we see building across the optical and AI infrastructure markets.

Our customer engagement is deepening our visibility, is improving and our competitive differentiation is strong. While we remain disciplined and thoughtful in our execution, we're confident that the groundwork laying out now will enable us for meaningful growth in the years to come.

With that, I turn the call back to Gary for our second quarter guidance. Gary?

Thank you, Morris. To reiterate a couple of key points from Morris's commentary, we are seeing a strong increase in our Indium Phosphide wafer demand related to AI and the ongoing data center upgrade cycle. Given the geopolitical complexity surrounding this market trend, our customer base is diversifying and expanding, and customers are placing longer-term orders and providing greater visibility into their needs.

With all of these positive market and AXT-specific growth drivers the most significant single factor to our growth in Q2 and beyond is the success and timing of getting export permits. Therefore, guiding for future revenue is somewhat tricky for us right now as we cannot predict future timing of permits or our success in obtaining them for any customer or individual order. But drawing on what we know and what we've experienced thus far in the export permitting process, we can offer the following insight into our expectations for Q2.

As of today, we have approximately $34 million in revenue that can be realized in Q2 across our substrate product lines and raw materials for which we either already have a permit to ship or for which an export permit is not required.

We have a high degree of confidence in recognizing this revenue in Q2. We could see upside, even significant upside to this number in Q2, should we receive permits for additional orders for which we have the inventory to support. But we do not want to, but we do want to stress that the timing for permit issuance is not predictable nor in our control and doesn't align with our quarterly reporting.

We continue to focus on gross margin improvement. Further improvement depends on a number of factors, including total revenue as it relates to revenue mix by product, absorption of fixed costs and our ability to continue to drive better manufacturing efficiency.

With regards to OpEx, we expect that it will be approximately $9.3 million in Q2 on a non-GAAP basis and approximately $10 million on a GAAP basis. With these factors in mind, we expect to achieve profitability on both a GAAP and non-GAAP basis in Q2. We believe our non-GAAP net income will be in the range of $0.06 to $0.08, and our GAAP net income will be in the range of $0.05 to $0.07. We estimate share count for Q2 will be approximately 63.5 million shares.

Okay. This concludes our prepared comments. We'd be glad to answer your questions now. Tracy?

[Operator Instructions] Your first question comes from the line of Tim Savageaux with Northland Securities.

Congrats on the step-up in backlog and the strong guidance for next quarter in Indium Phosphide. I guess my first question, you mentioned backlog and customer forecast at record levels, and we certainly saw that with $100 million in backlog. With regards to long-term capacity planning with customers, are you at the point of coming to any sort of long-term supply agreements with various customers? And if so, what's the kind of, what sort of timing might you'd expect on that?

Yes. Thanks, Tim. Yes, we are talking to a number of customers right now on long-term supply agreements as we build our capacity out and try and understand where their demand is going. Nearly all of the larger customers in this space are talking to long term, talking about long-term supply agreements with us. And we expect to come to resolution with some of those in the very near future.

Great. And just following up on that, sorry, go ahead.

No. I was just saying it's your turn now Tim.

Okay. Just an update, you mentioned last quarter that you were developing some relationships with Tier 1 customers or Tier 1 suppliers who hadn't necessarily been close relationships or customers over time. I wonder if we can get an update on that. And I have one more follow-up after that.

Yes. Thanks. That's going really well, actually. We've got qualification wafers in with a lot of customers, and we're finding paths and avenues to get wafers into a lot of these Tier 1 customers. We've, as we see this market grow, there's a lot of opportunity for us. And we've said in the past that we've really been focused on these next-generation technology products, that require high-quality material that, frankly, only AXT can build and can supply.

And of course, with emerging supply chain constraints with Indium Phosphide, we are in the strongest position to grow capacity. So, we're qualifying and we're supplying wafers to a lot of new Tier 1 customers in this field. So, it's exciting times for us.

Yes. I want to add one point because I think Tim, you are the friend's soldier, you're talking to them. But from my perspective, I started to hear, let's say, three months ago was some of the Tier 1 customers. But now I'm starting to hear even add on to it, is the end hyperscalers we're hearing. In other words, the customers' customer, the end users are also interested in seeing how we develop the supply chain guarantee for their growth plan.

Yes, that's correct, Morris. That's a good point. So, there's been a lot of press releases out about long-term supply agreements into our customers from the hyperscalers and from the hardware companies. And there's been a lot of encouragement from those hyperscalers and hardware companies for their suppliers to enter into long-term supply agreements with AXT.

So that is actually driving a lot of the discussions, I think, that we're having on long-term supply agreements. And of course, it's given us a lot more visibility into what the market demands are at the hyperscaler side of things, and how that trickles down to demands for AXT. It also gives us a lot of visibility into technical demands as we move forward into high-end lasers and detectors in these new products.

Great. And that makes sense and maybe somewhat related to those discussions. I'd be interested in an update on what you're seeing in terms of pricing for Indium Phosphide substrates.

Yes, that's a good question. Again, thanks, Tim. So what we are doing is we are raising some of our prices. We're seeing some recent pricing increase in raw materials and specifically with Indium.

So we're having conversations with our customers to align our costs and maintain gross margins, maintain or grow gross margins. We're also starting to globalize or we've been globalizing our pricing. Obviously, certain geographical regions have been more aggressive in the past on price targets, especially when we're looking at the lower-end markets such as GPON. So we're starting to globalize our markets so that it is more standardized across those geographical regions.

Well, let me add on to that. I think, nevertheless, I think the pricing opportunity for us, I believe, is also the fact we're migrating more towards larger size. As you know, some of the smaller size, they are more traditional they are more price sensitive and they have more competitors who can fill that shoes. But when you get to 4-inch and 6-inch and then as well as higher specification requirement, then we can really demand that's where our product shines.

Your next question comes from the line of Matt Bryson with Wedbush Securities.

Great results. I just wanted to hone in on the gross margins a bit. Obviously, you saw a pretty big uptick in Q1. I'm not quite getting to the peak you had in Q2, in Q2, I'm not quite getting the peak back in the COVID time frame, but I'm getting pretty close. I guess, could you talk a little bit about how much of that is higher utilization levels versus how much of it is increased pricing and whether my math is roughly accurate?

Well, for Q1, there is some that's a result of increased pricing, but it's the primary drivers are the traditional 2 drivers that we highlight. One is volume is up and the other is the mix is rich towards Indium Phosphide. As a matter of fact, if you look at it percentage-wise, Indium Phosphide was just a tad north over 50% of total revenue. So it's really helped in that. The pricing effects are being put in place, but we'll see them, we'll see the impact from your viewpoint, Matt, for your eyes, your eyes shall see that later this year.

Just Q2, Gary, if I said that the gross margins are coming in roughly around 40%. Is that in the ballpark? And again, can you just talk to how much that's mix versus utilization versus pricing?

Yes. I don't have your forecast in front of me, but I think that's too aggressive. And you know us, we like to take, be a little bit more conservative. So we're definitely going to be crossing the 30% threshold, which we said for several years, if we can get to $30 million in revenue and have a good mix, and we could be above 30% in gross margin. So, but I would. I think you're probably, it's up to you, but I'd encourage you to maybe knock that down a bit. We can talk about it maybe later. So, but having said that, we're both on the right direction. Gross margins should go up, and we feel very confident that they will. And how fast and what we calculate is to be determined. But all the indicators are exactly what I've been saying for many years now, and the mix is rich for Indium Phosphide and the volume is up. And so it's a unique sort of transition for us that's inside the company, we're very pleased, very pleased.

Yes. I do want to add other point about this. I mean, obviously, Gary, you own the gross margin calculation. But I would argue the supply chain strategy will start to shine. I always say the AXT is like a choo-choo train with the locomotive. We're chugging along when we are accelerating all the box behind us, such as our JinMei, BoYu, which makes our crucibles, high-purity materials, et cetera, they all going to chug along with us. When we slow down, of course, they will crush against us. But right now, it's a good time. We're chugging along very strongly. So you're going to see their contribution to our ability to make profit will grow, too.

And then just my one follow-up is, I noticed going back to the last filing that you've got export licenses, I think, for every geography, except for the U.S. just any more thoughts on getting licenses for shipping in the U.S.? And how important is that in terms of being able to fully utilize that additional capacity you're bringing on?

I don't think we're giving up the United States. No. It's still pending.

Right. We're still Matt, we're still being encouraged to apply for export permits for U.S. customers, both in the U.S. and in other global regions. So at the moment, obviously, we're getting, are getting permits pretty readily for U.S. customers based in other global regions. But that as Morris said, that doesn't mean that we are completely stopping any work on trying to obtain permits for the U.S. We've been commented or we've been contacted by the Ministry of Commerce in China on a number of U.S. applications right now to submit more data that gives us an encouraging sign that they're still looking at U.S.-based permits, and there's still a possibility to get a permit for the U.S. in the near future. as I say. So we are definitely looking at that avenue.

And in the meantime, we are supplying wafers globally to other regions as well. So this is a very global supply chain, and it's a very global market. And I think we're taking advantage of all the avenues that we can.

Yes. And I didn't mean to intimate that you weren't getting, going to get a U.S. permit or were still working on it. I just want an update and that was an update.

[Operator Instructions]. Your next question comes from the line of Charles Shi with Needham.

I want to ask you more about the capacity and the capacity build plan here. I think your last COVID high for Indium Phosphide quarterly record was $17.7 million. That was achieved in second quarter 2022. You're basically implying you're going to be at or above that level in this coming Q2.

But I recall back in 2022, you probably also built above that 17.7 because back then, you thought you would have Indium Phosphide demand from the premium electronics company for smartphone applications.

So, I want to ask you this, what's the max factory output for your existing factory today? How utilized is the existing factory? And what's the expected capacity factory output once you add the next 2 factories. I mean, I think that's something you talked about after the follow-on offering. And if you can provide any color on the numbers, that would be great.

Yes. I usually take the last digit out. We usually say our highest Indium Phosphide revenue per quarter was $17 million. You have a very good memory, okay? And we did say in Q4, we said we have increased our capacity by about 25% in Q4 of 2025. And in 2026, we're going to double that, okay? So in my calculation, our own capacity planning, we think we're going to get about $35 million per quarter capability by the end of 2026. Okay? But don't forget, that's the end of 2026.

In other words, the capacity are increasing every month, every time as we talk about, look, I mean, in the next quarter, our Indium Phosphide revenue is going to be up and beyond the $17 million per quarter.

Will be a new record.

Will be a new record in Q2, okay? But the other capacity we also mentioned about is that we are acquiring other piece of land near our existing factory in Beijing right now, which is, we're in the process of negotiating buying the land and doing the design, and we're probably going to start building it. But because it's a greenfield, it will probably take us about a year, maybe 1.5 years to complete that expansion, okay?

However, our capacity expansion is sort of in stages. For instance, our, sometimes the clean room is the most critical because if you don't have the clean room, you don't have no space to put in your machine. So that's very digital. And, but some of the crystal growth capacity is more incremental, okay? So right now, the clean room capacity is greater, much greater than our crystal growth capacity.

So as we speak now, we're increasing our capacity sort of gradually. So, but I think in the next year or so, once the greenfield is up for construction, then I think it would be more digital to expand our clean capacity. I mean, do you have anything to add, Tim?

Yes. So I just want to add a little bit. Morris talked about doubling our capacity to a rate of $35 million per quarter in Indium Phosphide by the end of this year. Remember, that's in a brownfield site that was once a crystal growth facility used for Gallium Arsenide. And as we relocated gallium arsenide, we've been able to move into that. So we've been extremely fortunate that we're in a position, I think, that nobody else in the Indium Phosphide world is in that we can double our capacity so quickly.

Looking into the next growth, as Morris just mentioned, we are, we're acquiring a facility, which is right next door to us. Again, extremely fortunate., building is already there. And that allows us to double yet again. So by the time we've completed that expansion, which should be by the end of 2027, maybe early 2028, we should be at the region, of somewhere in the region of $65 million to $70 million of capacity per quarter. So that really takes us to the type of capacity that we're expecting to see in our existing locations.

And then as Morris mentioned again in his call or in his script, as we talked earlier, we're now looking at where we need to go from here. So we're looking at other opportunities and other ways to expand beyond that probably in a greenfield site somewhere else.

Yes. And that's for 2028.

Correct.

So let me be a bit more specific as sort of the detailed guide, but $17 million, which we've already achieved. By the end of this year, we'll be at $35 million.

Per quarter.

So that's times 4, $135, $140 million.

Per quarter?

Per year. Yes. So we'll have that capacity at the end of 2026 to do $140 million.

But you cannot do that because the capacity is continually increasing. You can't use that.

It's analog. A year later, it will be $280 million. So it's double, double, double.

Maybe a follow-up on the capacity expansion, right? I think this is not like I come up with a question, but investors do ask this question. When you think about your capacity expansion, why can't you do the China plus 1 type of strategy like many companies in the global economic, electronic supply chain, like maybe you should continue to build in China, to satisfy China demand, but can you build outside of China, maybe to supply to the rest of the world?

And I know this is more, this is an easier step than done. There are policy reasons that may stop you from doing that. But is there anything you think from a business perspective that is preventing you from doing that and why?

Well, there's certainly a lot of opportunity, both within China and outside of China for us to consider that. And as I said just now, we're looking to build more capacity in 2028 and beyond, which is going to be meaningful capacity expansion in 2028 and beyond. And as part of that plan, we are working closely with our customer base to understand the long-term requirements and aligning the plans globally, right?

So, our recent capital raise will be fundamental to expanding as we enter this next growth plan, which could include more capacity within China, potentially also with capacity outside of China.

Yes. So, I do want to add one point. I know you don't want me to say this, okay? I tell you, important thing to answer to investors is that adding capacity versus able to deliver wafers is two different things. You're going to hear a lot of people who are going to say, I'm going to add capacity. Look, Indium Phosphide, I tell you, it's not easy. And one question a lot of investors are asking me is you're going to double your capacity, why don't you triple or quadruple. Our need is 10x. It's not easy.

Morris, that's a really good point. And that's really why our focus on the next two years has been focused on Beijing and increasing the capacity on our existing Beijing, Tongmei site. That is the minimum risk that we can absolutely take to get wafers out, not just to increase capacity.

Well, not only that, it's also for the good of our customers. Their demand is so aggressive, the better way or the guaranteed way to satisfy that capacity and we're stretching, we're working very hard to answer their demand is to do now, okay? Do we have other plans? You bet. We do. We're stepping up, don't forget.

Yes. Thanks, Morris. Like I said, easier said than done. Just felt like that's a question so common that I have to ask. So maybe last question. You talked about 6-inch versus maybe 4-inch or below. What's the shipment or maybe shipment is a bad metric here. But what's in the backlog that you are seeing the mix between 6-inch and 4-inch or the low right now? And I want to get some thoughts around that. And if you can, also, we're kind of curious about the mix between Iron-doped and Sulfur-doped is for laser, the other is for photodetector. I want to get some thoughts on what's the mix expected mix within that $100 million plus backlog. That's the last question.

So iron doped is coming up big time. we used to see about 10:1 in favor of sulfur doped prior to this. Right now, I would say the mix, especially the large diameter, it's almost like iron is 40%, 60%, correct, Tim?

Yes. So when we look at backlog and we look at customer demand over the next few quarters into next year and beyond, what we can see is there's still a lot of 3-inch out there, specifically for the laser. So sulfur doped is still going strong on 3-inch. There is a transition to 4-inch on N-type material for the laser, whereas the high-speed detector, frankly, has pretty much all transitioned to 4-inch already. So we're seeing still a lot of 3-inch coming along, a lot of transitioning over to 4-inch. And as we look into the future, of course, 6-inch is incredibly important. And there's a lot of interest and a lot of opportunity out there for 6-inch.

But I'll say at this moment in time, a lot of the production that we're seeing and a lot of our capacity that we're seeing is still focused on the 3-inch and the 4-inch with a longer-term plan to transition to 6-inch within the next probably year or so.

Yes. The signals are obviously very strong. A lot of customers are telling us, can we get more 4-inch, okay? And 6-inch is actually a little bit more out, but people are warning us it's coming, it's coming, okay? But 4-inch is real. And I would say the ratio for inch 3-inch and 6-inch right now is maybe 4:1 in favor of 3-inch. And I think going out in about 6 months to a year, it could be, I mean, as far as wafer number is concerned, it becomes probably 2:1 in favor of 3-inch. 3-inch is still the majority, the larger, the numbers. But because 4-inch is actually at a lower number right now, so it's going to grow very rapidly in the next coming quarters.

Your next question comes from the line of Richard Shannon with Craig-Hallum.

A couple of questions here. I'd love to understand how do we think about the CapEx requirements here for these capacity builds? You talked about a brownfield one this year that's doubling. And then a greenfield one, I think that's going to happen in '28 or maybe in '27, maybe going into '28 here that's more greenfield. Wonder if you could give us some numbers or at least some statistics to think about what that's going to require and over what period of time?

Well, for this year, it's mostly adding high-tech growth equipment for crystal growth. So furnaces, some back-end stuff for polishers. But as Tim and Morris have said, we have an existing footprint. That's one reason we think we have an advantage. Our current Indium Phosphide Crystal growth site has room for more furnaces.

And as Tim explained, we're repurposing our gallium arsenide crystal growth that was in Beijing for even more. So this year, it's compared to the future years, it's probably going to be $35 million in CapEx, maybe $30 million, maybe $40 million, somewhere in that range. And to be honest, we'll spend as much as we can, as fast as we can because we're uniquely positioned to be able to add capacity quickly. So next year. Let's see. I think as we, it's depending on which things we're talking about. Tim, do you want to, you've got split up in your paperwork there, but.

Yes. I think as we go into next year and we look at building out this facility next door to us, obviously, buying a new facility, doubling our capacity again there and also building some capacity through our supply chains as well. I think we're looking somewhere in the region of about $100 million or so.

And then beyond that, if we were to build a greenfield site somewhere else, I think we're looking at somewhere in the region of $220 million to $250 million, depending on obviously what capacity we put in that greenfield site. But again, I think if we're putting a meaningful capacity there, you're looking at $200-plus million.

Fair enough. I want to ask on your Indium Phosphide business here by geography. You made a couple of interesting comments here. Last call, you said China was going to grow about 60% this first quarter, and then you said it actually was up 100%, if I caught you correctly, it's going to double again here in the second quarter. What kind of percentage of Indium Phosphide business in the second quarter is China going to be?

That's a great question. So I think we're seeing a lot of growth in China, and it's not just because we're seeing data center growth in China, but we're seeing China enter the global supply chain market for optical transceivers and potentially co-packaged optics as we go forward.

So again, remember, this is fully globalized and a lot of those transceiver companies that manufacture their transceivers within China are driving to a Chinese supply chain of laser diodes and photodetectors.

So in Q2, we estimate that the Chinese demand is probably about 30% of the overall Indium Phosphide global market demand that we're seeing. And we're seeing that increasing through, certainly through Q3 and Q4 as well. So as we get into Q4, it could even be as high as something pushing up to 40% share of the total Indium Phosphide market.

That is helpful, Tim. And I'll ask one last question and jump out of line here, and that's on the topic of gross margins. Gary, you've talked about in the past here, with hoping to get to 35% with kind of an upside goal of looking at 40%.

But when you're talking about the pretty strong mix shift towards Indium Phosphide here and even about price increases here, I would imagine you maybe help us think about whether that could go higher at some point in time. I'm not asking for anytime soon, but are you looking for kind of a ceiling of gross margins that get us above that 40% level?

Well, internally, as a management team, we're definitely going to be targeting something that begins at the 4%, but it's far out. We don't know yet. And so, I'd still stick with my sense that somewhere in the 35% range is very, very reasonable. But that's for the outside world. It's a safe arrival landing point. But that doesn't mean that we're satisfied with it, and we think we can do better, but let's, we need to get it further down the road and prove that first.

There are no further questions at this time. I will now turn the call back to Leslie Green, Investor Relations at AXT for closing remarks.

Thank you, Tracy, and thank you all for participating in our conference call. We will be participating in the B. Riley Securities 2026 Annual Investor Conference and the Craig-Hallum Institutional Conference in May as well as the Northland Virtual Conference in June. We hope to see many of you there. And as always, feel free to contact us if you'd like to set up a call. We look forward to speaking with you all in the near future. Thanks.

This concludes today's call. Thank you for attending. You may now disconnect.

Good afternoon, everyone, and welcome to AXT's Fourth Quarter 2025 Financial Conference Call. Leading the call today is Dr. Morris Young, Chief Executive Officer; and Gary Fischer, Chief Financial Officer. In addition, Tim Bettles, VP of Business Development, will be participating in the Q&A portion of the call. My name is Audra, and I will be your coordinator today. I would now like to turn the call over to Leslie Green, Investor Relations for AXT. Please go ahead.

Thank you, Audra, and good afternoon, everyone. Before we begin, I would like to remind you that during the course of this conference call, including comments made in response to your questions, will provide projections or make other forward-looking statements regarding, among other things, the future financial performance of the company, market conditions and trends, emerging applications using chips or devices fabricated on our substrates, our product mix, global economic and political conditions, including trade tariffs and import and export restrictions, ability to obtain China export permits, timing of receipt of export permits our plan to list our subsidiary, Tongmei in China, our ability to increase orders in succeeding quarters to control costs and expenses, to improve manufacturing yields and efficiencies or to utilize our manufacturing capacity.

We wish to caution you that such statements deal with future events, are based on management's current expectations and are subject to risks and uncertainties that could cause actual results or events to differ materially. In addition to the matters just listed, these uncertainties and risks include, but are not limited to, the financial performance of our partially owned supply chain companies increased environmental regulations in China.

In addition to the factors just mentioned or that may be discussed in this call, we refer you to the company's periodic reports filed with the Securities and Exchange Commission. These are available online by link from our website and contain additional information on risk factors that could cause actual results to differ materially from our current expectations. This conference call will be available on our website through February 19, 2027.

Also, I want to note that shortly following the close of market today, we issued a press release reporting financial results for the fourth quarter of 2025. This information is available on the Investor Relations portion of our website. I would now like to turn the call over to Gary Fischer for a review of our fourth quarter results. Gary?

Thank you, Leslie, and good afternoon to everyone. Revenue for the fourth quarter of 2025 was $23.0 million compared with $28.0 million in the third quarter of 2025 and $25.1 million in the fourth quarter of 2024.

To break down our Q4 '25 revenue for you by product category, indium phosphide was $8.0 million, primarily from data center applications, gallium arsenide was $7.0 million, germanium substrates were $231,000. Finally, revenue from our consolidated raw material joint venture companies in Q4 was $7.6 million.

In the fourth quarter of 2025, revenue from Asia Pacific was 81.5%, and Europe was 17.5% and North America was 1%. The top 5 customers generated approximately 22.6% of total revenue and no customers were over the 10% level.

Non-GAAP gross margin in the fourth quarter was 21.5%. For comparison, we reported 22.6% gross margin in Q3 of '25 and 18.0% gross margin in Q4 of last year. For those who prefer to track results on a GAAP basis, gross margin in the fourth quarter was 20% -- 20.9% compared with 22.3% in Q3 of 2025 and 17.6% in Q4 of 2024. We continue to be highly focused on driving continued improvement, including further recovery in Q1.

Moving to operating expenses. Our total non-GAAP operating expense in Q4 was $7.8 million, compared with $6.5 million in Q3 of 2025. As a reminder, Q3 included some favorable adjustments in R&D that brought our OpEx down to a lower-than-normal level. Non-GAAP OpEx in Q4 of '24 was $9.8 million.

On a GAAP basis, total operating expense in Q4 '25 was $8.8 million compared to $7.3 million in Q3 and $10.6 million in Q4 of 2024. Our non-GAAP operating loss in the fourth quarter of 2025 was $2.6 million compared with the non-GAAP operating loss in Q3 of 2025 of $384,000 and a non-GAAP operating loss of $5.4 million in Q4 of 2024. For reference, our GAAP operating line for the fourth quarter of 2025 was a loss of $3.8 million compared with an operating loss of $1.1 million in Q3 of 2025 and an operating loss of $6.2 million in Q4 of 2024.

Nonoperating other income and expense and other items below the operating line for the fourth quarter of 2025 was a net gain of $285,000. The details can be seen in the P&L included in our press release today.

For Q4 of 2025, we had a non-GAAP net loss of $2.6 million or $0.06 per share compared with a non-GAAP net loss of $1.2 million or $0.02 per share in the third quarter of 2025. Non-GAAP net loss in Q4 2024 was $4.2 million or $0.10 per share.

On a GAAP basis, net loss in Q4 was $3.6 million or $0.08 per share. By comparison, net loss was $1.9 million or $0.04 per share in the third quarter of 2025. GAAP net loss in Q4 of 2024 was $5.1 million or $0.12 per share. Weighted basic shares outstanding for the quarter was $44.7 million.

Cash, cash equivalents and investments increased by $97.2 million to $128.4 million as of December 31. This is primarily the result of our public offering of common stock, which closed on December 30 and generated approximately $93.9 million. By comparison, at September 30, cash was $31.2 million, accounts receivable decreased in the quarter by $2.6 million.

Depreciation and amortization in the fourth quarter was $2.3 million. Total stock comp was $1.3 million. Net inventory was up by approximately $4 million in the fourth quarter to $81.7 million. This continues to be a focus for us, and we expect to bring it down in coming quarters.

This concludes our discussion or comments about the quarterly financials, turning to our plan to list our subsidiary, Tongmei in China on the STAR Market in Shanghai, we remain very interested in completing the IPO, particularly in light of the rapidly evolving AI infrastructure build-out in China, which is fueling increased China-based demand for indium phosphide substrates. We've continued to keep our IPO application current, and Tongmei remains in process as part of much of a more selective and smaller group of prospective listings than a few years ago.

Though the current geopolitical environment is dynamic, Tongmei is considered a Chinese company and continues to be regarded in China as a good IPO candidate. We will keep you informed of any updates. With that, I'll now turn the call over to Dr. Morris Young for a review of our business and markets. Morris?

Thank you, Gary.

[indiscernible] interruption, this is the operator. We have lost our speakers. Give me one moment to reconnect.

[Technical Difficulty]

Let me start on the beginning again, just in case I missed part of it.

We were disappointed that we didn't receive as many export permits in Q4 as we had hoped. Based on the average processing time we had seen up to that point in October. The good news is now that we have received permits in Q1 and we are in a stronger position today than we were at the same time in the prior quarter.

Gary will take you through our full quarter guidance in a few minutes. But we do expect to achieve sequential growth in revenue in Q1, driven primarily by growth in indium phosphide for data center build-out for AI.

We're also very pleased to note that we are seeing a work expansion of our customer base for indium phosphide. We're beginning to support leading customers in the optical space that we have not -- we have limited exposure to prior to this time. This includes Tier 1 laser manufacturers and optical transceiver module makers, both in China and around the world. We're excited to be able to demonstrate the technological advantage of our low EPD wafers as the market moves to optical devices with higher speed and greater sophistication for both scale-up and scale-out applications.

In total, our backlog for indium phosphide wafers have reached a new high of over $60 million. As we mentioned last quarter, customers are planning for longer lead time by placing longer-term motors and giving us more visibility into their expected demand. As many of you know, the supply chain for optical transceiver is quite complex and highly globalized. We believe this geographical interdependence is providing both opportunity and incentives for the ecosystem to work together in new ways to solve global supply chain shortages.

Beyond pluggable receivers, we are seeing a very large developing market for co-packaged optics for both scale-up and scale-out applications. We're actively engaging in discussions with our customers about their technical and timing requirements and believe this could be -- represent yet other inflection point in our business developing in late 2027 and beyond.

For geographic demand perspective, the massive AI infrastructure build-out and planned CapEx spending by cloud services and AI platform providers in the United States is the primary driver for [ EML ] and silicon photonic-based optical transceivers. We believe that today, our materials are being used in multiple U.S. hyperscale and we expect that end customers use will continue to broaden.

In China, the data center build-out is early in its ramp. But we are seeing rapid growth as China moves to accelerate its data center expansion and AI capabilities. Our revenue related to the data center market in China are expected to grow by more than 60% in Q1 over Q4. Highlighting both increased investment in these Tier 1 data centers as well as the strong desire for Chinese domestic suppliers to secure local stores at every level of the AI infrastructure supply chain.

Given the strong demand environment, it is important to note that AXT is well positioned to handle increased demand for indium phosphide wafers. Since we have last reported to you in October, we have already added approximately 25% more capacity, and we are on track with our current plan to double our capacity from Q4 2025 level by the end of this year.

Beyond our current plan for capacity expansion, we're working closely with our customer base to understand their long-term requirements and to align our plans globally. Our recent capital raise will be fundamental to our future expansion as we enter our next significant phase of growth.

A major focus of this expansion will be an increased investment in our 6-inch indium phosphide product, and we are excited to work with our customers to meet the rigorous requirements of next-generation and co-location space devices.

Now turning to gallium arsenide. We continue to see demand for semi-insulating wafers for wireless RF devices and believe that we have strong opportunity for market share expansion gated primarily by our ability to obtain export permits. In we saw an uptake in semiconducting wafers for both industrial laser applications and data center laser applications. VCSEL lasers a data center for data center applications typically do not require a lot of gallium oxide material. As the device are small, so they don't move the middle much as a growth driver for us.

However, we are seeing increased demand for VCSEL for autonomous vehicles in China -- Chinese automobile market, which is currently expanding rapidly. High-growth expenses, in addition to our watching we are watching with interest an emerging application in robotics for VCSELs that increase the physician and dexterity of a modern robotic hand.

Counter the VCSEL used in data center applications, machine mission VCSELs tend to be very large and use more gallium oxide substrate. They also require high-quality material which we are very well positioned to supply. Again, demand is more today, primarily China-based and covers a diverse set of customers but the breadth of use case and the development is very exciting.

Finally, our raw material business is -- was up in Q4 with growth from our subsidiary volume, which manufactures [ pBN ] crucibles used in manufacturing of indium phosphide crucibles. In addition, we're pleased to report that our subsidiary, JinMei, has begun to refine high-quality indium, which gave us -- now direct control of a guaranteed supply of yet the other critical material for our indium phosphide substrates.

Globally, there continues to be a greater awareness of the importance of various materials, and we are ahead of the curve in developing our unique integrated supply chain.

In closing, this is a very dynamic and exciting time for our company as we enter into 2026 we're a fundamental supplier to the multiyear optical build-out in the AI infrastructure market. We have a broadening customer base of Tier 1 companies and a strong balance sheet to support our continued business expansion.

And with growing backlog the receipt of indium phosphide and gallium arsenide export license remains the single most significant gating factor for our growth. As such, we are highly focused on ensuring that we are proactive, organized and disciplined about managing the process on behalf of our customers.

We also know that we must be laser-focused on running our business with the greatest efficiency. This includes our continued effort to drive gross margin improvement, OpEx discipline and inventory reduction.

With strong ongoing market trends fueling the data center upgrade cycle we believe that we have tremendous opportunity in 2026 to drive meaningful growth in our business and return to profitability.

I would like to personally thank our employees for their dedication and tireless efforts during this singular moment in AXT history and while we also like to express my sincere gratitude to our customers, partners and shareholders for their ongoing support and believe that in the future, we are building together. We look forward to reporting to you on our progress. And with that, I will turn the call back to Gary for our fiscal quarter guidance.

Thank you, Morris. To reiterate a couple of key points for Morris' commentary, we are seeing a strong increase in our indium phosphide wafer demand related to AI and the ongoing data center upgrade cycle. .

Given the geopolitical complexity surrounding this market trend, our customer base is diversifying and expanding and customers are placing longer-term orders and providing greater visibility into their needs. With all of these positive market and AXT specific growth drivers, the most significant single factor to our growth in Q1 and beyond is the success and timing of getting export permits.

Therefore, Guiding for the future is somewhat tricky for us right now as we cannot predict future timing of permits or a success in obtaining them for any customer or individual order. But drawing on what we know and what we've experienced thus far in the export permitting process, we can offer the following insight to our expectations for Q1.

As of today, we have approximately $26 million in revenue that can be realized in Q1 across our substrate product lines and raw materials, for which we either have already have a permit to ship or for which an export permit is not required. We have a high degree of confidence in recognizing this revenue in Q1.

We could see significant upside to this number in Q1, should we receive more permits for additional orders between now and the end of the quarter. But we want to stress that as we experienced in Q4, the timing for permit issuance is not predictable nor in our control and doesn't necessarily align with our quarterly reporting.

As Morris mentioned, we continue to focus strongly on gross margin. Further improvement depends on a number of factors, including total revenue as it relates to absorption of fixed costs, revenue mix by product and our ability to continue to drive better manufacturing efficiency.

With regards to OpEx, we expect that it will remain at approximately $9.0 million in Q1. With these factors in mind, we believe our non-GAAP net loss will be in the range of $0.02 to $0.04 and GAAP net loss will be in the range of $0.04 to $0.06. This represents substantial year-over-year progress towards our return to profitability. We estimate share count in Q1 will be approximately 53.2 million shares. Okay. This concludes our prepared comments. We'd be glad to answer your questions now. Audra? Operator?

[Operator Instructions] We'll go first to Richard Shannon at Craig-Hallum.

Gary, I'm going to do a quick request to give me the revenue number you gave for the quarter. It got my line got garbled here. I heard about '26 that you believe you can get highly likely to get. Was there a number to the upside there? Apologies for needed to ask this.

We normally give you guys a range, but we discussed before the call today that we're very, very confident at the '26 number. We did say just a moment ago that we believe we could go higher if we get more permits, but we it wouldn't even -- it could even be more than just a normal range, which we usually have a $2 million or $3 million range for you guys.

Well, let me try to add on to this point. That is -- our manufacturing are doing the manufacturing as if we can get a permit. So there is a lot of these so-called [indiscernible] finished goods, or finish good staging in our clean room ready to be shipped if we can get an actual permit.

Yes. We are building to forecast and to the backlog, whether or not -- we're not building to permits. We're not waiting until we get a permit and then say, okay, let's get going. And so it's building and we're enthusiastic, we're excited and of course, yes, we're a little bit frustrated because it would be pretty big numbers that we can get some more of these permits. And we think that we will. We can comment more on this call, but we're hanging in there. We're not discouraging giving up.

Okay. I appreciate understanding your approach to the guidance and it certainly makes a lot of sense in this environment. Let me ask about the licensing process here. Last quarter, you said it was about a business day or a 3-month process here. And obviously, that didn't turn out as we saw from your pre-announcement, which is unfortunate, but we all know how governments can work from time to time here. Have any new insights as to how they're working here? And I guess, are there any permits that are being rejected that you don't think should be? Just more insights here on this licensing process.

Yes, I can answer that one. So this process is not transparent at all. And we're seeing quite a lot of variability. We started off in the end of Q3 by saying it was looking like we're seeing a fairly consistent 60 business day process cycle. We're now seeing a lot more variability as we go through there. And as I say, there's just no transparency to that.

It's reasonable to assume that there's geopolitics playing into this as well. It's really hard to determine what and why. And it's difficult, therefore, to figure out which permits are coming in on time, which are taking longer.

I'll answer the second question as well, which you asked whether there had been any permits that have been denied and why? We have actually received a couple of denials with the instructions that we can resubmit that application with more information. So this is the first time we've actually received denials on permits and we're not totally sure why, again, no transparency to this. We don't see any particular reason why any of these permits should not be approved. And it's a process that we're just working through.

So these permits that have been denied, we've already resubmitted with [ MOCOM ] and we're hopeful that we continue to talk to [ MOCOM ] and they will get reviewed quickly and could potentially turn around fairly quickly. I could even make an impact on Q1 numbers if we can get a quick turnaround on them.

And what does MOCOM stand for?

That's a Ministry of Commerce in China.

Yes. So let me add what optic viewpoint, the comment about the team just give you. That is -- although there is a denial of an application, but they come with a specific instruction how to strengthen the application, which we think is a good indication. In other words ,if they really want to deny this is one of -- they can just let it sit there. I mean the fact that they want more information about -- actually, I think it's a fixable permit application we have. And that means, hey, they are taking a very serious look at it. And hopefully, that will turn to be approved.

Okay. That is helpful. Second question here is kind of the backlog here and also following up on more of your comments about customers booking further out. So we went from a backlog of [ 49 to above 60 ] here, and you also commented that people or customers are ordering further out. Could you suggest how far out they're going right now? And also, how far out are you hearing forecasts from these customers as well?

Yes. Let me see how to answer that. Actually, let me first answer my part of the question, and I will turn it over to Tim. Well -- and the reason why that Tim really works with customers hearing what their demand is actually Well, the interesting comment we have was that we have important meetings with our customers this week, and they're telling us -- Tim at least in two occasions, people are saying, gee, our demand forecast increases every week. So that's the kind of level. I think -- I mean we know it is tight and we know it's going up. But I think people are upsizing their demand, and they're telling us what they want to do, whether they're going to go to [indiscernible], how much they want to switch on 3 to 4 and 4 to 6, okay?

And as far as how much inventory they are building, I think that depends upon customers. Some of the customers, we suspect they're buying into the inventory. But they also tell us I'm going to take it all in consider whereas others, I think they are telling us the real demand in the quarter because I think as of now, we cannot deliver enough of their demand. So they are giving us longer lead time to give us more incentive to build up the expansion plan and build the capacity for them. And also, I think the other thing is, Tim, you want to comment on long-term commitment that you're talking to a customer about?

Yes, I definitely -- I'll comment a little bit on that, and I'll also comment a little bit more on backlog and what we're seeing from this. So a lot of this backlog, remember, is scaled up based on the permit dynamics, right? So the permit dynamics, once we receive a permit to export material we have a 6-month window to export. So a lot of backlog is built at the moment that permit comes in, we have a 6-month window maximum to deliver. And in many cases, that window, the window of which the customers are looking to receive material is a lot shorter than 6 months. So really and truthfully, this is all being gated by permits, as we mentioned during the discussion. .

In terms of what we're seeing in build-out for inventory, I think at this moment in time, people would like to keep more inventory. But as Morris mentioned, just about everybody we're talking to is telling us that the demand is growing literally on a weekly basis. So we just see the numbers expanding and expanding over and over.

Now turning to forecast and what kind of visibility we have we are definitely talking about long-term supply agreements with a number of customers right now. And we're planning our business according to those long-term supply agreements.

We're seeing forecast out beyond 2030 for many of these customers, but of course, as I've just said, those numbers are increasing on a week-by-week basis. So it's difficult to keep track of things, but people are talking about minimum demand requirements. Moving forward for at least the next 2 to 3 years, given its forecast out beyond 2030. So all in all, I think this backlog is real. It's achievable, and it's kind of being limited by our permits at the moment.

Makes sense. I'll ask one more question and jump out of line here, guys. This is on capacity additions. Just a few kind of multipart question here. I think I heard you say you're going to double your capacity from the end of '25 to the end of '26. If you could verify that? And if so, can you help us understand what level of CapEx is going to be requiring?

And then looking beyond that, and Tim, you just mentioned forecast going out beyond 2030, which is interesting to hear, how much more capacity beyond that could you need? Let our minds wonder about what kind of scale an opportunity you're thinking about here?

Yes. I think we just said we have increased our capacity by 25%. And now, and we do expect to double our capacity by the end of this year, okay? And how much budget would we need? It could be about $30 million, and that is sort of on the low end in a way because the first phase of the expansion, which is doubling our capacity mainly use brownfield. In other words, existing Tongmei facility, we already have a clean room available. We have the building there already and power supply and water. So I think that budget is lower.

Looking beyond 2026, we are looking at possibly doubling it again in 2027. And that budget is lining somewhere around $100 million to $150 million depends upon how we want to build it because then we are talking about a greenfield. We need building, we need clean we need power, et cetera.

We'll move next to Tim Savageaux at Northland Securities.

Let's continue with that capacity discussion, but maybe try to put some numbers around it. If I look at where you've peaked historically, and I think we're talking exclusively about indium phosphide here, that's getting up towards $20 million a quarter in substrate revenue and I imagine your capacity is now slightly above that, given the increase you talked about in Q4.

I guess question one, is that reasonable? And should we expect you to exit calendar '26 with revenue capacity roughly double those levels? And would you anticipate having a demand of to fulfill that at that time where you're building maybe a little bit ahead.

Let me first answer the question. The -- I think we calculated -- I think it's approximately $35 million a quarter by the end of the year, run rate, okay? It could be a little bit more given the price environment is dynamic as we -- the cost of indium are going up.

And can we use up all this capacity? I think looking at the backlog, we can certainly do, but the problem is the gating factor is the permit.

I'll add something in here as well. Irrespective of permits, we mentioned we are seeing growth in this business in China as well. and look at it quarter-to-quarter, Q4 was probably double revenue in China than Q1 in 2025. And we would expect -- we're looking at potentially doubling again through 2026. So we're definitely seeing growth there in China that warrants expansion as well as growth outside of China where we would need permits for.

I agree with you, but then I don't want to minimize the importance of outside of China. But I think the AI growth budget we're seeing is really fueling the demand for indium phosphide substrates.

Yes, Tim, this is Gary. And I still speak conservative, but -- and I am. But I'm not sure you guys are getting the point is that every customer is worried about getting enough for their needs. There's a general concern. The meetings we've had this week, we're not meeting with the purchasing manager. We're meeting with CEOs and general managers. They all want to talk to Morris about capacity and about future growth.

So there's a phenomenon going on here that all of us -- it's unusual for -- no matter what we do for our jobs, including the analysts, is very unique and unusual situation. I mean I've been around the block a few times and Morris has, and this is very, very unusual. And it's actually intense. We're excited but we're scrambling. We're scrambling. And I don't see any into it near term. This is -- people are telling us that their demand is going to be going up 3, 4 or 5x over the next 4 or 5 years. And there's not how many suppliers are there. You know the answers to that, too, and we're one of them.

Yes. I think let me add to that. I think the investor usually asked the CEO, the toughest question is what keeps you up at night? I think what's keeping up at night is calculating how we're going to expand that capacity, how we're going to get that product to our customers and how to develop the technology that a customer wants. I mean it's very exciting, but it's also very training. We need to be very much aware of what the customer wants and satisfy the demand.

Fortunately, we have recent experience at adding capacity. What was almost about 10 years ago, when we learned that we needed to get gallium arsenide moved out of Beijing. And so we had 2017, '18 kind of time period where we did add capacity from green grass fields. So we have some strength here but it's going to tax us even though we are experienced.

Yes. I do want to give the analyst point to ask the question, but I think we're talking to each other.

We're excited, yes.

But I think it is a very good point. I think we -- prior to this, we probably overspend because the IPO preparation, we actually expand from one facility to three facilities. But now I think we're looking at a great demand for indium phosphide which I think it's really meeting our challenge. And I think Gary is right. We are very well positioned to meet that demand. I think we are probably the best suited to increase capacity and also because the virtual integration we have in terms of supply chain, and we're in control of a lot of other material, which cooking to supply if indium phosphide continue to grow like what we are talking about, and we have plans for that as well.

A good example is our subsidiary, JinMei. JinMei make the indium phosphide, poly, for Tongmei. So we have -- our supply chain is supporting this growth process. Next question, Tim?

I'm a little bit afraid now. But you actually you highlighted in the release even the increased presence with some big Tier 1 customers. I guess in the commentary, you mentioned maybe some in China, but elsewhere.

In terms of what's going on there? Are we talking about orders with major new customers qualification? Any specific programs? And I'm not sure these two comments were related or not. But I'll ask if they were with regard to your increased investment in 6-inch indium phosphide, if you can maybe cover both of those points. Appreciate it.

Yes. I'll take a stab at that one, Tim. So yes, we are gaining more traction with customers, as we've said on previous calls as well that we've not been so prolific in. So we are gaining design in, we're gaining qualifications on existing products as well as new products as we move forward. And the customers are looking to expand on their demand for indium phosphide. .

As Morris mentioned in the call, there's already been a big move from 3-inch to 4-inch, so we've spent a lot of time and effort on scaling up our 4-inch business. And we're also seeing a lot of interest now. And of course, we all know one customer that's really driving 6-inch demand. So we're really taking 6-inch very, very seriously. And we're expanding -- as we expand capacity both now and are doubling capacity through '26 and beyond, we're looking at adding significant 6-inch capacity in there during that expansion. And we're just plowing through the numbers right now to see what we need to drive 6-inch and how we scale 6-inch compared to 3, 4 and more of the traditional wafer sizes.

Yes. I think one part of it perhaps is the cooperative effort. Usually, when your customers don't go to me, they probably talk to the sales guy and give us orders. But I think now the dynamics is such that we sort of need to interact more to make sure that we're putting the right amount of attention both in terms of development and capacity expansion to where they need it, okay. And then virtually also to convince us, this is the right investment we should have. Is that right?

That's correct. We're also getting a lot more customer buy-in with commitments, [ NRE ] purchase orders to drive that business forward as well.

We'll move next to Matt Bryson at Wedbush Securities.

Just can you talk a little bit about what might have been unfettered demand or shipments in Q4 or what you might be guiding to if you won't restricted by permits?

If we're not restricted by permits, then the basic question is our ability to manufacture high volume because there's no issue about demand or backlog. So the variable that you're really focusing on is manufacturing capability.

Yes. I don't know whether the customers are telling us more demand than we can deliver. But I think we definitely have more orders than we can now. As we add the capacity, we're counting on who we can supply to, but of course, there's other bidding factor, which is the permits.

Got it. So I mean hypothetically, assuming you could be manufacturing around $20 million [indiscernible] fund, you could ship it all if you could get permit?

Yes, correct. And that will -- we expect it to increase it to about $35 million a quarter by the end of the year. And that we are making sure every point along the supply chain receives equal attention in terms of poly in terms of crucible furnaces, et cetera, et cetera.

Understood. And then so you're completely confident that of that $35 million in capacity, if you can bring it on and you can get permits that come to end of this year, you could possibly be shipping $30 million, $35 million in orders. I guess is there -- are there any customer commitments or [ LTAs ] or anything else that kind of solidifies that demand?

What's that turn you're seeing? LP?

We've got a lot of purchase order in that right now, and we're going through long-term agreements. Long-term agreements, I think in terms of locking up capacity are easy and we're talking to customers to lock that capacity up.

The gating factor, and we've reiterated this a lot today and previously, gating factor with long-term agreements is how much can we actually ship out of the country? What can we get permits for. So we're trying to address that through LTAs. But for sure, we can definitely cover this kind of revenue volume with purchase orders and LTAs.

And then, Gary, I think the last one I have is for you. if you get to those numbers in terms of shipments for indium phosphide, so $30 million plus, can you give us some of the parameters we should be thinking about in terms of gross margins, what are the puts and takes? And hypothetically would be able to get back to the kind of COVID era highs you're reporting back in 2021, 2022?

Yes. I mean we always like indium phosphide. And if you have to pick of our three substrate products that you want to see go through the ceiling in terms of volume and demand is the right one for us. I think getting somewhere at $40 million a quarter in aggregate, not just indium phosphide, but we should be getting hopefully somewhere close to 35% gross margin.

I will add another point. I always describe AXT, it's a fairly unique company because we I describe our substrate business as the locomotive engine in the front, but we have a lot of cars in the back following us, such as indium, such as phosphorus, quartz, [ pBN ] in crucibles furnaces we make. So if our business is good, we're pulling these guys along, and that should help us. So what you're seeing -- I'm more optimistic and Gary said, I think that 35% is the normal substrate business. If we can pull those guys along that should help us even further.

Yes. Yes. Our internal goal is higher, Matt, but -- so that we don't overstate expectations from for your community, we want to be a little bit cautious. So we're very optimistic -- so it's pretty exciting.

And just one quick follow-up. The shift to 4-inch and 6-inch does that change the parameters on a gross margin perspective?

I think normally the larger the size we go, the better the margin we will get. On the other hand, I'll be more cautious about 6 inches. We are still a little bit in development stage. So I think initially, looking at lower margin, but looking for ways to compensate that. Right Tim?

Right. Product mix is still very much geared towards 3-inch and 4-inch at the moment. And as Morris says, we're running 6-inch up. It is still a bit of a development project at the moment. It will be growing through this year. But again, remember, as we do that, 3-inch and 4-inch are still a big percentage of our business.

And that concludes our Q&A session. I will now turn the conference back over to Leslie Green for closing remarks.

Thank you, Audra, and thank you all for participating in our conference call. We will be participating virtually in the Loop Capital Conference in March and hope to see many of you there. As always, feel free to contact us if you would like to set up a call, and we look forward to speaking with you soon.

And this concludes today's conference call. Thank you for your participation. You may now disconnect.

Good afternoon, everyone, and welcome to AXT's Third Quarter 2025 Financial Conference Call. Leading the call today is Dr. Morris Young, Chief Executive Officer; and Gary Fischer, Chief Financial Officer. In addition, Tim Bettles, Vice President of Business Development, will be participating in the Q&A portion of the call. My name is Kelvin, and I will be your coordinator today. I would now like to turn the call over to Leslie Green, Investor Relations for AXT. Please go ahead.

Thank you, Kelvin, and good afternoon, everyone. Before we begin, I would like to remind you that during the course of this conference call, including comments made in response to your questions, we will provide projections or make other forward-looking statements regarding, among other things, the future financial performance of the company, market conditions and trends, emerging applications using chips or devices fabricated on our substrates, our product mix, global economic and political conditions, including trade tariffs and import and export restrictions, ability to obtain China export permits, the timing of receipt of export permits, ability to increase orders in succeeding quarters to control costs and expenses, to improve manufacturing yields and efficiencies or to utilize our manufacturing capacity. We wish to caution you that such statements deal with future events are based on management's current expectations and are subject to risks and uncertainties that could cause actual events or results to differ materially.

In addition to the matters just listed, these uncertainties and risks include, but are not limited to, the financial performance of our partially owned supply chain companies and increased environmental regulations in China. In addition to the factors just mentioned or that may be mentioned in this call, we refer you to the company's periodic reports filed with the Securities and Exchange Commission. These are available online by link from our website and contain additional information on risk factors that could cause actual results to differ materially from our current expectations.

This conference call will be available on our website at axt.com through October 30, 2026. I also want to note that shortly following the close of market today, we issued a press release reporting financial results for the third quarter of 2025. This information is available on the Investor Relations portion of our website at axt.com.

I would now like to turn the call over to Gary Fischer for a review of our third quarter 2025 results. Gary?

Thank you, Leslie, and good afternoon to everyone. Revenue for the third quarter of 2025 was $28.0 million compared with $18.0 million in the second quarter of 2025 and $23.6 million in the third quarter of 2024. To break down our Q3 '25 revenue for you by product category, indium phosphide was $13.1 million, primarily from data center and PON applications. Gallium arsenide was $7.5 million, germanium substrates were $640,000 and revenue from our consolidated raw material joint venture companies in Q3 was $6.7 million. In the third quarter of 2025, revenue from Asia Pacific was 87%, Europe was 12% and North America was 1%. The top 5 customers generated approximately 45.2% of total revenue and 2 customers were over the 10% level. Non-GAAP gross margin in the third quarter improved substantially to 22.4%, reflecting improved product mix and higher volume to absorb overhead.

For comparison, we reported 8.2% gross margin in Q2 of 2025 and a 24.3% gross margin in Q3 of 2024 last year. For those who prefer to track results on a GAAP basis, gross margin in the third quarter was 22.3% compared with 8.0% in Q2 of 2025 and 24.0% in Q3 of last year. We continue to be highly focused on driving continued improvement, including further recovery in Q4. Moving to operating expenses. Given the difficult climate, we've been working hard to hold down OpEx. In addition, we had some favorable adjustments in R&D in Q3 that brought our OpEx down to a lower-than-normal level. These will not carry over into Q4. Therefore, our total non-GAAP operating expense in Q3 was $6.7 million compared with $7.6 million in Q2 and $8.3 million in Q3 of 2024.

On a GAAP basis, total OpEx in Q3 was $7.3 million compared with $8.2 million in Q2 and $9.1 million in Q3 of 2024. Our non-GAAP operating loss for the third quarter of 2025 improved substantially to $384,000 compared with the non-GAAP operating loss in Q2 of 2025 of $6.1 million. and a non-GAAP operating loss of $2.6 million in Q3 of 2024. For reference, our GAAP operating line for the third quarter of 2025 was a loss of $1.1 million compared with an operating loss of $6.7 million in Q2 and an operating loss of $3.4 million last year in Q3. Nonoperating other income and expense and other items below the operating line for the third quarter of 2025 was a net loss of $46,000. The details can be seen in the P&L included in our press release today.

For Q3 2025, we had a non-GAAP net loss of $1.2 million or $0.03 per share compared with a non-GAAP net loss of $6.4 million or $0.15 per share in the second quarter of 2025. Non-GAAP net loss in Q3 of 2024 was $2.1 million or $0.05 per share. On a GAAP basis, net loss in Q3 was $1.9 million or $0.04 per share. By comparison, net loss was $7.0 million or $0.16 per share in the second quarter of 2025. GAAP net loss in Q3 of 2024 was $2.9 million or $0.07 per share. The weighted average basic shares outstanding for Q3 2025 was 43.8 million shares.

Cash and cash equivalents and investments decreased by $3.9 million to $31.2 million as of September 30. By comparison, at June 30, it was $35.1 million. Accounts receivable increased by $11 million, so the delta in cash is explained in working capital. Depreciation and amortization in the third quarter was $2.3 million. Total stock comp was $0.7 million. Net inventory was down by approximately $2.4 million in the third quarter to $77.7 million. This continues to be a focus, and we expect to bring it down further in quarters to come. This concludes the discussion of our quarterly financial results.

Turning to our plan to list our subsidiary, Tongmei in China on the STAR Market in Shanghai. We've continued to keep our IPO application current. Tongmei remains in process as a part of a much more selective and smaller group of prospective listings than a few years ago. Although the current geopolitical environment is dynamic, Tongmei is considered a Chinese company and continues to be regarded in China as a good IPO candidate. We will keep you informed of any updates.

With that, I'll now turn it over to Dr. Morris Young for a review of our business and markets. Morris?

Thank you, Gary. This has been a very eventful quarter for AXT as we are seeing a strong uptick in indium phosphide demand from data center applications globally and as our industry and our customers adapt a new normal within a rapidly changing environment. In Q3, our revenue grew 56% sequentially and 18% year-over-year. Within this, our indium phosphide revenue grew to our highest level since 2022 as we were successful in obtaining export permits for a number of significant indium phosphide orders throughout the quarter. I'm very proud of the diligence our team and grateful for the partnership of our customers in working through the export control permitting process. Our current experience is that our indium phosphide permits are taking approximately 60 business days or approximately 3 months to be processed by China's Ministry of Commerce. This is a bit longer than our initial expectations, but customers are adapting to the requirements and are adjusting their ordering patterns to give us more visibility and longer lead times.

I should also note that the Golden Week holiday at the beginning of October in China will likely increase the average permit processing time by a week or so in Q4. The tremendous growth in demand for indium phosphide-based lasers and detectors for high-speed optical connectivity, coupled with our successful obtaining export permits on behalf of our customers are driving a strong increase in our indium phosphide order backlog, which as of today is more than $49 million and growing. Our established customers are planning for longer lead times by placing longer-term orders and giving us more visibility into their expected demand. We are also seeing active engagement with several Tier 1 -- new Tier 1 customers to qualify our material into their supply chains for the first time in many years. This include leading optical transceiver module makers, both in China and around the globe.

As many of you know, the supply chain for optical transceiver is quite complex and highly globalized. We believe this geographic interdependence is providing both opportunities and incentives for the ecosystem to work together in new ways to solve global supply chain shortages. For a geographic demand perspective, the massive AI infrastructure build-out and the planned CapEx spending by cloud services and AI platform providers in the United States is the primary driver for EML and silicon photonics-based optical transceivers. We believe that today, our materials are being used in multiple U.S. hyperscalers, and we expect that end customers' use will continue to broaden. In China, the data center build-out is early in its ramp, but there is a strong desire for domestic suppliers at every level of the supply chain, and we believe over the next 12 to 18 months, we will see healthy growth in the China data center market.

Data center expansion in China is quickly overtaking PON as the leading application in China for our indium phosphide substrates. Given the strong demand environment it is important to note that AXT is well positioned to handle increased demand. We have ample manufacturing capacity in place today, and we can also significantly increase our output by current level, and we can also add capacity quickly as needed. We also have a demonstrated ability to supply very low EPD wafers in volume that meet the vigorous requirements of next-generation EML and silicon photonics-based devices.

Now turning to gallium arsenide. Our revenue grew more than 20% from the prior quarter. The biggest driver was semi-insulating wafers for wireless RF devices, which remains a focused application for us. Industrial laser applications were about flat from Q2, and we saw an uptick in semiconductor wafers for data center laser applications. However, VCSEL lasers don't typically require a lot of gallium arsenide material, so they don't move the needle much as a growth driver. But they do require high-quality material, which we are well positioned to supply. In germanium substrates, our sales declined by about $1 million in Q3.

The germanium substrate market was very poor gross margin potential today. And while our material performed well in the solar cell applications as we supply, gross margin constraint dis-incentivize us to pursue many opportunities. In addition, certain customers prefer to source substrates outside of China. As such, we do not expect growth in germanium substrates in Q4. Finally, our raw material business in Q3 was consistent with the prior quarter and it was solidly profitable within a stable pricing market. We expect the same for Q4. Globally, there continues to be a greater awareness of the importance of earth materials, and we are ahead of the curve in developing this unique integrated supply chain.

In closing, this is a highly active time for our business. The receipt of indium phosphide and gallium arsenide export permits remains the single most significant gating factor for our growth. As such, we are highly focused on ensuring that we are proactive, organized and disciplined about managing the process on behalf of our customers. We also know that we must be laser-focused on running our business with the greatest efficiency. This includes our continued effort to drive gross margin improvement, OpEx discipline and inventory reduction. With strong ongoing market trends fueling the data center upgrade cycles, we believe we have tremendous opportunity in 2026 to drive meaningful growth in our business and a return to profitability. We look forward to reporting to you on our growth, on our progress.

With that, I will turn the call back to Gary for our fourth quarter guidance. Gary?

Thank you, Morris. To reiterate a couple of key points from Morris's commentary, we are seeing a strong increase in our indium phosphide wafer demand related to AI and the ongoing data center upgrade cycle. Given the geopolitical complexities surrounding this market trend, customer behaviors in our space are changing to allow for longer substrate lead times. our customers are placing longer-term orders and providing greater visibility into their needs. As such, our indium phosphide backlog has grown to $49 million and is the largest we've ever had in our history.

Further, we are actively engaging with new customers today that we have not had business with an opportunity for some time. With all of these positive market and AXT-specific growth drivers, the most significant gating factor in our growth in Q4 and beyond is the success and timing of getting export permits. Therefore, guiding for the future is somewhat tricky for us right now as we cannot predict future timing of permits or our success in obtaining them for any customer or individual order. But drawing on what we know and what we've experienced thus far in the export permitting process, we can offer the following insight into our expectations for Q4.

As of today, we have approximately $20 million in revenue that can be realized in Q4 across our substrate product lines and raw materials for which we either already have a permit to ship or for which an export permit is not required because it ships within China. We have a high degree of confidence in recognizing this revenue in Q4. In addition, we believe there's an incremental $7 million to $10 million in indium phosphide and gallium arsenide backlog, which is currently in our manufacturing process for which we believe we may be able to ship in Q4 if we are awarded permits. Of course, timing of permits is not within our control, but we believe we are in a similar or slightly better position in terms of customer order backlog and permit submissions than we were at the same point in the prior quarter.

As such, with that as a background, we believe we have the capability to achieve revenue in the range of $27 million to $30 million in Q4, subject to the caveats I just mentioned. This takes into consideration approximately flat sequential revenue contribution from germanium substrates and raw materials with incremental growth in Q4 likely coming from indium phosphide and gallium arsenide substrates. As Morris mentioned, we continue to focus strongly on gross margin, we made significant gains in Q3 and continue to work on our manufacturing efficiency. Further improvement in Q4 depends on a number of factors, including total revenue as it relates to the absorption of fixed costs, revenue mix by product and our ability to continue to drive better manufacturing efficiency.

With regards to OpEx, we expect that it will increase to approximately $9 million as a result of some incremental end of the year adjustments and a return to a more normalized level. With these factors in mind, we believe our non-GAAP net loss will be in the range of $0.01 to $0.03, and our GAAP net loss will be in the range of $0.03 to $0.05. This represents substantial year-over-year progress towards our return to profitability. We estimate the share count for Q4 will be approximately 43.8 million shares. And okay, this concludes our prepared comments.

We'll be glad to answer your questions now. Operator, Kelvin?

[Operator Instructions] your first question comes from the line of Charles Shi of Needham & Co.

Maurice, Gary, congrats on receiving the licenses, the permits shipping $8 million additional revenue in the quarter, and congrats again on the $49 million backlog. That was an exciting number to hear. I really want to get back to this point on the customer behavior change, like are they placing longer-term orders. But I think if I hear you correctly, some of those customers may not necessarily have the permits at this point and still proceeded to place the orders with you, a pretty significant amount of orders with you. Can you kind of talk through what exactly is driving that behavior? And what do you think the export permits, currently -- the current ones you already have, are there time limits to that? Are there like the volume limits to that? And what could be your best prediction going forward from here, the customer behavior can continue to evolve?

Yes. Thank you, Charles. So we have, as you say, $49 million backlog. That includes customers that have previously received permits and customers that are still in the permit phase for the first permit as we go through. Everybody that has previously received a permit has typically received subsequent permits from there. So there's a lot of confidence in getting further permits as we move forward through this. So people are placing orders into that backlog with the understanding that the confidence levels of receiving permits are high, especially for indium phosphide. So as we look forward and as we look at that backlog, all of the orders that we've received and put into backlog have permit applications in place so far. And we manage that backlog and those permit applications, and we manage the manufacturing process so that we can combine the expected permit approval time with the finishing of the product. So our lead time to ship the product after receiving the permit is very low.

Yes. So maybe I can add another point. I hear Charles is asking why? Is there any relationship with customers giving us a lot more order, a longer order lead time? Is it because we have a permit process? I think that is true. People realizing instead of just in time, they want to give us a long lead time to submit the permit application so that we can ship this product to them in time. Is that a part of the question, Charles?

Yes. I think maybe a better way to help us understand what the permit to the size of the orders, how much long term the orders is going to be? Maybe you can shed some light on, let's say, the order currently on average cover is it like 1-year demand, 2-year demand, 3-year demand? What do you see there? Like how long does the order you have in the backlog covers what customers demand?

Right. Okay. Understood. Thanks, Charles. So the permit, we apply for a permit and it can be for multiple shipments, number of shipments up to 12. This is the important part. The permit only lasts 6 months. So everything has to be shipped within 6 months of receiving the permit.

Yes. And the other point is this, our customers are telling us, we give you this order, if you get the permit and if you can manufacture it, you can ship it tomorrow.

So when Tim mentioned up to 12, that means 12 line items. Every PO needs a separate permit. So if you put each line item on a separate PO, then we need 12 permits. It's complicated as they say in the show business.

It is, it is. So maybe I ask another question on profitability. So when you were at this revenue level in the high 20s, going back a few years, you probably have a gross margin somewhere in the high 20s or even low 30s percent and you would have a non-GAAP EPS in the positive territory. But I think, Gary, if I hear you right, I think you're still expecting some gap -- some non-GAAP loss in the coming quarter. And wonder if there's anything in your cost structure that's a little bit different now versus back then? And how do we get back to like the similar profitability level at the similar revenue run rate back in the, let's say, go back -- only go back 2 or 3 years, yes.

Yes. We expect it to be asked that question to us. So -- and this is something that we talk about internally. So as I like to say to ourselves and to analysts and investors, in our business model, it's never one single dial. It's not like one thing we can focus on, and we have to focus on 2 to 4 things to sort of move the needle in the right direction. In this regard, one of the things we need to get improvement on is gross margin. And that's primarily a result of mix, which is going in our favor right now and also efficiencies on the line. So we have -- I'm actually encouraged to be able to say this because it's pretty much in our control. And we've done better than we're doing right now. But this is -- it is common in manufacturing businesses to have some cycles. And so I think we can work on that and focus on it and get improvement. That's probably the biggest one. I think we could get a bit more help from our joint venture companies. I expect that to improve in the coming quarters as well. But those are the two things that come to mind.

So Charles, maybe I can answer part of the other question. I think the deadliest thing in manufacture, I think analysts should ask is, is your ASP dropping, okay? I think we can say except with the low end on the 2-inch indium phosphide, most of our ASPs are holding very firm. In fact, some of the ASP for our high-end low EPD indium phosphide substrate, the ASP is increasing. okay? So I think we can surely stop that worry. I mean we have some other efficiency issues such as loading factors, germanium is perhaps not making a whole lot of money for us because the pricing pressure is very strong. But the main focus on indium phosphide, the pricing is firm and the demand is high.

I think maybe one last question before I jump back into the queue would be the indium phosphide demand you are seeing today, how much of that is from the overseas customers that would need a permit versus domestic Chinese customers? And if I recall correctly, I remember that the indium phosphide was primarily shipped to outside of China previously. How much of the domestic development today maybe has led to a little bit more of a domestic shipment of the indium phosphide. If you can kind of paint a little bit of picture to us of how things have been evolving, that would be great.

Well, actually, indium phosphide business is very globally connected. A lot of our substrates are shipped to, let's say, Taiwan to put EPON and ship back to United States to make a device and ship back to China to make a transceiver and then ship back to U.S. data centers. So I think -- but our direct customer in China is roughly, I would say, 40%. But the great AI opportunity definitely is the big increase is in the AI data center in the United States.

And I think I can add to that as well. If you look at our financials from Q2 versus Q3, you can see that the indium phosphide in Q2 was about $3.5 million, and that's increased to about $13 million in Q3. So that kind of gives you an idea of what the incremental is and all of that incremental has come from outside of China.

Your next question comes from the line of Richard Shannon of Craig-Hallum.

I'll offer congrats on a wonderful quarter. Great to see. So congrats to the entire team for making that happen here. Let's start with the first question here on the indium phosphide backlog. I just want to understand the dynamics here. Maybe if you can help us understand a few things here. What was the backlog a quarter ago? And then how far out are customers ordering here? I would imagine, given one of the prior answers you're talking about a permit allows you to ship for 6 months that they're probably going out 6 months here, but just want to get a sense of what this looks like and how it's changed.

Yes. So as Morris has previously said, those permits do last 6 months, but most of our customers are asking to ship as soon as we can. So that backlog -- once we have a permit, we can ship that backlog as quickly as we can manufacture, to be perfectly honest. So in terms of our backlog last quarter, we've got more than double the backlog in -- as we speak today than we had last quarter. So that continues to grow. And as we said in the conference call, we're seeing more and more new opportunities coming. So that backlog is growing daily as we speak.

Maybe I can chime in a bit. I think as CEO, I take care a lot of this China development, engineering and manufacturing and also my duty is to push the IPO process in China. But recently, I got putting more and more to talk to customers of indium phosphide because they cannot get enough material, they call my sales guys and the sales guy says, well, you got to come and visit the customers to calm them down. How are we opening up the opportunity to supply indium phosphide customers. So I got a lot this very good warm receptions from our customers and sometimes the customers' customer and also the end user. So in other words, the epi growers, the device makers as well as the CPU, GPU makers.

In fact, I got the message from our customers, especially globally, not especially, they all told us that we are a very important supplier of indium phosphide. Secondly, they all told me there's a great, great opportunity to increase the demand in the near future. Obviously, they are all anxious to know what are we going to do to ease the pain of getting the permits to export material. And lastly, quite a few customers told us they start to appreciate the better EPD or the better quality of indium phosphide material we supply. In fact, one customer told me that now every die count and using our substrates, they can make better die yield on their lasers or detectors. So I think that's a very warming information for me and also telling us that indium phosphide, the paradigm -- there's a paradigm shift because of the global increased demand for AI connectivity in optical transceivers. And what's the other words, CPOs?

CPOs.

I'll start to learn that one.

Okay. That is helpful. I'm going to explore a couple of different angles on the dynamic here. So I think one of the things that investors will be worried about or cognizant of here is customers understanding the geopolitical dynamics, as you referenced in your prepared remarks and worried about the door shutting here at any point, very well could be ordering well above what their normal rates of consumption would be and building some level of inventory. To what degree do you see that behavior anywhere here in the backlog build? And what are the limits to your shipping faster? Are you near full utilization in your indium phosphide today?

So let me start by answering the dynamic question. I think the fact of the matter is that people are building inventory levels so that they have inventory on hand. But I don't think this is a onetime build-out because they're concerned. This is a multiyear cycle. So the demand today that we're seeing is real, and you can see evidence of that all up and down the supply chain for optical transceivers, right? I want to really just look at some of the CapEx spending messaging that was given from U.S. hyperscalers on their earnings calls yesterday, right? So everybody is talking about CapEx moving faster and growth in dollars getting noticeably larger as we go through financial year '26.

So there's definitely growth going on here at the hyperscale level, and we're seeing that come into here. We're also seeing longer-term discussions on indium phosphide for CPO, both on scale up and scale across now. So the demand is there. The demand is real. And of course, people are building backlog -- or sorry, people are building inventory levels, but those inventory levels will continue to grow. So we don't see this as a one-and-done shot.

Yes. So let me add on to another point. Yesterday, Tim and I were in the valley visiting a few actually customers' customers. They are asking me what can they help in terms of financially, in terms of customer relationship to ensure that indium phosphide will be supplied. In other words, they are telling me there's a tsunami coming, okay? I just don't know how big the tsunami is because the normal rate, let's say, if it is one foot wave, then tsunami is only 10 feet. It's not that big. But if the normal wave is already 5 feet, then that's going to be very significant. So we're going to get that information soon. But I think the demand from what I hear is enormous. And don't forget, Richard, we are 40% of the indium phosphide supply chain, and we have the best quality material.

By the way, tsunami was used by the customer that Morris and Tim were visiting. We're not making it up in our conference room. So I was struck to hear that word as a description of what's on the future. So...

Okay. Now let me ask one -- another question here looking on the other side of this dynamic here, which is you mentioned a number of engagements with customers you've not worked with ever or for a very long time here. I think, Morris, you've been talking about the very good EPD specs on indium phosphide for a few years at least. And we haven't heard you talk about new customers really much, if at all. And I know I've asked on this conference call a few times in the last few years on this topic. Why is it there all of a sudden coming to you now? It seems like it's a unique or, I guess, a coincidental timing to see a number of customers coming to you at this particular time. What's going on here and what's driving that?

Jeez, you're so smart. I mean you call me. But I tell you, I have a perfect answer to that. That is, first of all, I think with all these lasers getting bigger and bigger, the EPD is getting that much more because the larger the device, the chances of you hitting a EPD is higher. In fact, yesterday, I was told by one of the customers, how come you guys can make the EPD so low, right, Kim?

Right, right. And I think the market is maturing such as well. And the demands that our customers are being faced with, with increased demands, increased capacity, one of the customers said to us, every device is important. The yield of devices on a wafer has become so much more important today than it ever has been, both because of cost and capacity constraints within the fab. So people are turning to us because they get much higher device yields from our wafers.

Yes. That's the customer told us straight in the face, they wouldn't tell us because we would have to ask higher price.

Richard, a secondary factor subservient to what Morris and Tim just described is there is a concern among the customer base about capacity and capacity potential. They're sensing that there are shortages, and we are the best positioned currently with capacity and with the ability to respond quickly to add capacity.

Well, Gary, that was a perfect setup for my next question here, which is on a full run rate basis, hand-to-mouth basis here, what is your kind of maximum indium phosphide revenues per quarter here? And how long would it take you to get a new capacity? And what kind of CapEx commitment to grow it by, I don't know, say, 25%? How does that look like?

Well, we could double our capacity on indium phosphide in about 9 months' time. It would take us about my estimation is -- because this is not a greenfield. We got the clean room already. We got land already and all we need to do is add a few crystal balls. So my estimation is about $10 million to $15 million. But we need a signal, I'm getting it.

So let me answer the question on current capacity there, Richard. So it's a complex question because it depends on a number of factors relating to product mix and wafer size and inventory on hand and all that kind of stuff. But we estimate that current capacity is around about $20 million a quarter for indium phosphide with our current run rate and current capacity that we've got. You asked how quickly can we increase by 25%? Probably within about 3 months, we can increase by 25%. We do not need to build anything other than bring some more furnaces online.

And to add for us to double that, then we need 9 months. That will be [indiscernible].

Your next question comes from the line of Tim Savageaux with Northland Capital Markets.

Again, congrats on that backlog number. Believe it or not, I still have a few questions. And I guess the overall question is, guys, is doubling capacity, is that a tsunami? Or is that just good business?

That's a good question, but I think I'll be happily retiring when the capacity is double with all the better gross margin. I'm joking. I think it's a lot more than that. I think -- but one step at a time, I think if we can double that, and I think we have all the ability to increase our capacity, well, the easiest way is in China. But I think beyond that, we may want to consider building something.

Yes. U.S.-based capacity would make a lot of sense. And yes, I think just intuitively, a tsunami is like 5 to 10x. And I have heard numbers like that in the industry in terms of where demand is going to be. And it sounds like the tsunami referenced in particular, is that a specific kind of looking forward scale up, scale across comments, which is to say, I assume what you're seeing in terms of current demand is likely module-driven, might be some early CPO, you tell me. But in terms of the real big step function in capacity, is that discussion mostly CPO-based or scale-up type based?

Yes, that's absolutely right. So we are seeing growth right now. That is, we believe, in the pluggable market and probably will continue to be in the pluggable market for the next few years. But we are starting to have those discussions now about growth rates for CPO for scale up. And the tsunami, the 5, 10x that you talk about, that is a lot of that is coming from CPO for scale up.

Yes.

Sorry Morris, you were saying something.

No, I'd say yes. Yes, I like...

The question -- I'll add to the -- just trying to get a sense of this backlog. So you increased your backlog, you doubled it and shipped $13 million in material, which I think gives you a book-to-bill that's approaching 3, so that's not bad. But where would that kind of normally be? I guess -- and so maybe as opposed to go back to last quarter, let's go back to last year or just historically without export permits required, what kind of backlog would you normally have in terms of quarters of revenue or just straight up, where was that indium phosphide backlog Q3 '24.

Well, because we could be responsive to customer orders, we had a lot of turns business every quarter. So to be honest, we don't really -- in terms of me and Morris and Tim, we don't manage the company by looking at a book-to-bill. Well I have in other companies, but it's not very meaningful in this case. But -- so it's hard to say what it was because I don't have a piece of paper in front of me with that list because there's no such list.

Got it. Well, it sounds like it should be some fraction, maybe half or 1/3 of whatever your indium phosphide revenue was a year ago. Your backlog is tsunami. It's up 10x, right?

Yes. And again, tsunami was not -- I agree with you, tsunami is 5 to 10x. And I'll say again, that was not our words. That was the words from an end customer. So...

Yes. Okay. Last one for me. You mentioned two 10% customers in the quarter. And Morris, you talked about kind of industry structure, epi Tier 1 back to the U.S. But any color on whether you've got an integrated device maker in there? Is this just really focused on epiwafer suppliers or whether you might have a new 10% customer in there?

Tim?

Yes. We've -- so the 10% customers that we've got, we've been dealing with for a while. The new customers that we've got are integrators as well. We're dealing more and more with integrators and hardware customers.

Including GPU and CPU makers.

Exactly right. So we're dealing with -- directly with GPU, CPU hardware makers. We're dealing with pluggable makers. So we're having -- that's where really the visibility is coming from.

Yes. I would say in the past, we haven't had access to those people. But now they're calling us. They want to see us. So that's why we had better visibility.

Your next question comes from the line of Matt Bryson of Wedbush Securities.

This is going to sound a little bit like a complaint, but it's not a complaint. Just curious, so there's clearly a whole lot of demand out there. Your Japanese competitor has announced 2 capacity increases in the last, I think, 4 months, 3 months. Just curious, if you have all this backlog and your customers want more product faster, why wouldn't you be building and shipping to capacity next quarter or this quarter?

Well, all of our shipments, all of our ability to ship is based on permitting. So we -- as we've talked about plenty of times, we've got a large backlog now, and we can ship as -- we've been told by customers, we can ship as quickly as we possibly can. But we have to go through the permitting process. Now that permitting process, it takes 60 business days, which is approximately 3 months. And there is some opaqueness to that permitting process. So if we had a bunch of permits today, I'm sure we could ship an awful lot more of that backlog today. We've guided at $27 million to $30 million. If we got permits, could we ship more than that? Yes, we could. But we're basically running trend analysis on how long it takes to get permits and a probability analysis of what permits we're going to get, and that's where the guidance comes in.

By the way, we're not standing still on those orders that we are applying for permits, we are putting that into WIP. In other words, we are making it, and we're packaging it and waiting for the permits to be issued and then we can deliver right away.

Got it. Understood. So I mean, it comes down to the permits of the gating factor, but as hopefully, permit approvals continue to get across the line and lift, there's a path to achieving the levels of shipments that you were at a few years back during COVID? And then I guess, what's -- in terms of gross margins, obviously, when you're running back at close to full capacity back then, you had substantially higher gross margins. I guess what's key to getting gross margins back up? Is it predominantly utilization? Or were you benefiting back then from higher pricing? Can you just talk to kind of the dynamics around gross margins, where they can go to from here if you can get indium phosphide back up to full utilization?

Yes. Pricing is not really a big factor. I think the big factor is volume because it does carry more of the fixed assets in a proper way. And -- and then it's -- I'm confident we can return -- we're going to be over 30% because there's -- we can control that. So we need to improve the efficiencies on the line, but I already commented on that. So I see it going in that direction.

I think the most important factor is we got more -- we can utilize our indium phosphide line. I think that's the greatest opportunity we're facing now.

[Operator Instructions] there are no further questions at this time. And with that, I will turn the call back to Leslie Green for closing remarks. Please go ahead.

Thank you, everyone, for participating in our conference call. We will be participating in the Northland Virtual Conference in December and the Needham Growth Conference in January, and we hope to see many of you there. As always, feel free to reach out to any one of us if you would like to set up a call, and we look forward to speaking with you in the near future.

Ladies and gentlemen, this concludes today's call. We thank you for participating. You may now disconnect.

Good afternoon, everyone, and welcome to AXT Inc.'s Second Quarter 2025 Financial Conference Call.

Leading the call today is Dr. Morris Young, Chief Executive Officer; and Gary Fischer, Chief Financial Officer. In addition, Tim Bettles, VP of Business Development, will be participating in the Q&A portion of the call. My name is Audra, and I will be your coordinator today.

I would now like to turn the call over to Leslie Green, Investor Relations for AXT. Please go ahead.

Thank you, Audra, and good afternoon, everyone. Before we begin, I would like to remind you that during the course of this conference call, including comments made in response to your questions, we will provide projections or make other forward-looking statements regarding, among other things, the future financial performance of the company, market conditions and trends, emerging applications using chips or devices fabricated on our substrates, our product mix, global economic and political conditions, including trade tariffs and import and export restrictions, our ability to increase orders in succeeding quarters to control costs and expenses, to improve manufacturing yields and efficiencies of or to use -- or to utilize our manufacturing capacity.

We wish to caution you that such statements deal with future events are based on management's current expectations and are subject to risks and uncertainties that could cause actual results or events to differ materially. In addition to the matters just listed, these uncertainties and risks include, but are not limited to, the financial performance of our partially owned supply chain companies, increased environmental regulations in China and COVID-19 or other outbreaks of contagious disease.

In addition to the factors just mentioned or that may be discussed in this call, we refer you to the company's periodic reports filed with the Securities and Exchange Commission. These are available online by link from our website and contain additional information on risk factors that could cause actual results to differ materially from our current expectations.

This conference call will be available on our website at axt.com through July 31, 2026. I also want to note that shortly following the close of market today, we issued a press release reporting financial results for the second quarter of 2025. This information is available on the Investor Relations portion of our website at axt.com.

I would now like to turn the call over to Gary Fischer for a review of our second quarter 2025 results. Gary?

Thank you, Leslie, and good afternoon to everyone. Revenue for the second quarter of 2025 was $18.0 million, compared with $19.4 million in the first quarter of 2025 and $27.9 million in the second quarter of 2024.

To break down our Q2 2025 revenue for you by product category, indium phosphide was $3.6 million, primarily from PON and data center applications in China. Gallium arsenide was $6.2 million, germanium substrates were $1.5 million. Finally, revenue from our consolidated raw material joint venture companies in Q2 was $6.7 million.

In the second quarter of 2025, revenue from Asia Pacific was 90%, Europe was 9% and North America was 1%. The top 5 customers generated approximately 30.9% of total revenue and 1 customer was over the 10% level.

Non-GAAP gross margin in the second quarter was 8.2%, reflecting a solid improvement from the prior quarter. For comparison, we reported a negative 6.1% gross margin in Q1 and a 27.6% gross margin last year in Q2 of 2024.

For those who prefer to track results on a GAAP basis, gross margin in the second quarter was 8.0% compared with negative 6.4% in Q1 and 27.4% last year. We continue to be highly focused on driving continued improvement, including further recovery in Q3.

Moving to operating expenses. Given the difficult climate, we have been working hard to hold OpEx down. Total non-GAAP operating expense in Q2 was $7.6 million compared with $8.5 million in Q1 and $8.9 million in Q2 of 2024. On a GAAP basis, total operating expense in Q2 was $8.2 million compared with $9.0 million in Q1 and $9.5 million in Q2 of 2024.

Our non-GAAP operating loss for the second quarter of 2025 was $6.1 million, compared with a non-GAAP operating loss in Q1 of 2025 of $9.6 million and a non-GAAP operating loss of $1.2 million in Q2 of 2024.

For reference, our GAAP operating line for the second quarter of 2025 was a loss of $6.7 million, compared with an operating loss of $10.3 million in Q1 and an operating loss of $1.9 million in Q2 of 2024.

Nonoperating other income and expense and other items below the operating line for the second quarter of 2025 was a net loss of $0.4 million. The details can be seen in the P&L included in our press release today.

For Q2 of 2025, we had a non-GAAP net loss of $6.4 million or $0.15 per share compared with a non-GAAP net loss of $8.2 million or $0.19 per share in the first quarter of 2025. Non-GAAP net loss in Q2 of 2024 was $0.8 million or $0.02 per share.

On a GAAP basis, net loss in Q2 was $7.0 million or $0.16 per share. By comparison, net loss was $8.8 million or $0.20 per share in the first quarter of 2025. GAAP net loss in Q2 of 2024 was $1.5 million or $0.04 per share.

The weighted average basic shares outstanding in Q2 of 2025 was $43.7 million. Cash and cash equivalents and investments decreased by $3.1 million to $35.1 million as of June 30. By comparison, at March 31, it was $38.2 million.

Depreciation and amortization in the second quarter was $2.5 million. Total stock comp was $0.6 million. Net inventory was down by approximately $300,000 in the second quarter to $80.1 million. This continues to be a focus for us, and we expect to bring it down further in quarters to come. Okay. This concludes the presentation of our quarterly financial results.

Turning to our plan to list our subsidiary, Tongmei, in China on the STAR Market in Shanghai. We have continued to keep our IPO application current. Tongmei remains in process as part of a much more selective and smaller group of prospective listings than a few years ago. While we are not insensitive to the current geopolitical environment, Tongmei is considered a Chinese company and continues to be regarded in China as a good IPO candidate. We will keep you informed of any updates.

With that, I'll now turn the call over to Dr. Morris Young for a review of our business and markets. Morris?

Thank you, Gary, and thank you to our customers and investors for your ongoing support as we navigate this unique macroeconomic environment. As Gary mentioned, our substrate revenue increased in Q2 from the prior quarter, but the increase was less than we had expected as a result of longer processing time for gallium arsenide export permits, coupled with some sluggishness in the demand environment in China.

That said, we made good progress in driving recovery in our gross margins with a strong focus on our manufacturing process and efficiency. We also saw healthy growth in AI-related demand for indium phosphide substrate in China. And as a result of obtaining our first export license in June, we were able to ship initial orders of indium phosphide substrates to our customers outside of China.

Since export restrictions are top of our mind for our investors, I'd like to begin there with an update, and then we will discuss our key markets. As many of you know, the China government imposed trade restrictions on export of gallium arsenide in August of 2023 and our indium phosphide in February of 2025. These regulations explicitly seek to restrict the export of materials used for military applications and require that we file an export permit for every customer orders.

In our experience, we typically hear back on initial applications with 45 business days, and repeat applications are often processed faster. With that said, we found the permitting process in Q2 for gallium arsenide to be slower than we typically see over the last 2 years. The delays in Q2 resulted in our being able to ship less material outside of China than we had anticipated. About half of our revenue shortfall in Q2 was the result of this factor.

On a positive note, the pace of permits in the month of July has improved meaningfully, mostly on smaller orders, but this improvement is good news, and we do expect gallium arsenide revenue to grow sequentially.

We're pleased to be granted our first permit for indium phosphide in late June, and we were able to ship nearly $700,000 of material from -- for our non-China backlog in indium phosphide in Q2. Although, the process for indium phosphide has been a bit slower than we expected as well, we were -- we have received additional permits in July and expect to see more over the coming months.

Based on the pace so far, we're taking a conservative view of the timing of larger permits in Q3. As we have mentioned previously, we don't believe that any of our indium phosphide sales go into military applications. So, we feel we are in a good position to realize millions of dollars of sales backlog once we navigate the permit process.

While the recent geopolitical environment present a near-term headwind for our business, we are also taking advantage of some of the unique opportunities. The cloud and data center connectivity market in China is accelerating. And in an effort to promote innovation and reduce dependency on foreign suppliers, we're seeing a significant effort to develop domestic source of EML and silicon photonics-based lasers.

We estimate that China data center optical interconnect market is currently around 1/3 of the global market. However, most of the optical devices for these interconnects are sourced from outside of China and applications for indium phosphide substrate within China remain focused on PON business only.

Further, laser manufacturers in China are developing an appreciation for the critical benefit of low EPD material in high-speed interconnect devices. Both in the traditional PONs market and in the new data center market. As a result, our sales of indium phosphide within China are increasing.

In Q2, we nearly doubled our revenue for AI -- for indium phosphide within China and our revenue for AI-related applications in China, although are increasing, although the revenue base is small, and we expect to continue to grow in Q3. The TAM for data center market in China remains small at this moment, but we expect to see significant growth over the next few years. As the PON laser providers expand their portfolio to include EML and silicon photonic solutions.

Broadly speaking, we expect to grow our total indium phosphide revenue by 30% or more in Q3, as a result of growth in applications for PON, data center connectivity and various indium phosphide-based sensors.

Now turning to gallium arsenide. Demand in China was sluggish in Q2, and customers are taking a more cautious approach to ordering and holding inventory. Despite a lackluster environment, we were pleased to be able to grow our wireless business in China during the quarter with continued growth expected in Q3.

As we mentioned, there's a sizable opportunity in the wireless market for which our technology and products are well suited. During Q2, we took a more measured approach to market expansion and we were able to service a portion of the customer opportunity, while executing effectively at modestly higher production levels.

The adjustment we made in our approach, along with the strong focus from our manufacturing organization on yield and efficiency, allowed us to drive meaningful improvements in gallium arsenide gross margins, which contributed to our overall progress to our good margin recovery. And this will, should continue to be a top priority for us in the second half of the year.

Turning to germanium business. We saw growth in our revenue in Q2, driven by satellite solar cell applications in China. This market is highly price sensitive, and we continue to be very selective in the business opportunities we choose to support as the sharp rise in germanium raw material pricing in the last several years has severely constrained gross margins.

In addition, germanium substrates permits for sales outside of China have been difficult to obtain. Therefore, in Q3, we expect to see our sales come down again, and we may remain at lower level rate through the second half of the year.

With regard to our raw material joint ventures, our consolidated revenue in Q2 declined by approximately $1.6 million compared to Q1. The economic climate was one factor and the other factor relates to the mix of revenue from the 2 service model a customer choose for our germanium -- for their gallium purification process.

On a positive note, the pricing environment has been relatively stable. Globally, there continues to be a greater awareness of the importance of earth material, and we're ahead of the curve in developing this unique and integrated supply chain.

In summary, though the expert permit process has been slower than we would like to be, we are making progress against a backlog of more than $10 million in customer orders for gallium arsenide and indium phosphide substrates. We're also encouraged to see growth in strategic applications within China, including indium phosphide for AI-related data center connectivity and gallium arsenide for wireless devices.

With a strong focus on gross margin improvement across our product portfolio, we delivered meaningful recovery in Q2 and expect to continue our progress in the second half of the year. Our competitive positioning continues to be enhanced by superior product performance in key specifications such as low EPD, and we are working diligently to support the next-generation technology requirements of our global customer base.

With that said, I will now turn the call back to Gary for our third quarter guidance. Gary?

Thank you, Morris. In keeping with our comments today, we believe Q3 revenue will grow sequentially to be in the range of $19.0 million to $21.0 million. This guidance range includes a modest contribution from indium phosphide and gallium arsenide for our customers outside of China and only includes revenue for which we currently have permits.

Within China, we continue to optimize the emerging opportunities to grow our business in strategic applications for both indium phosphide and gallium arsenide. And finally, we expect our germanium revenue to be down and our raw material business to be approximately flat in Q3 from the prior quarter.

As Morris mentioned, we continue to focus strongly on gross margin improvement. In Q3, we expect our margins to improve again and to be in the low mid- to mid-teens. Based on this revenue range and gross margin improvement, we believe our non-GAAP net loss will be in the range of $0.11 to $0.13 and GAAP net loss will be in the range of $0.13 to $0.15. Share count will be approximately 43.8 million shares.

Okay. This concludes our prepared comments. We'd be glad now to answer any of your questions. Audra?

[Operator Instructions] We'll go first to Ross Cole at Needham & Company.

And it's great to hear that you're starting to get some of the permits, especially for indium phosphide. But I was wondering, given that there's still been a bit of a delay in the permitting, are you concerned about any potential share loss to customers if they're continuing to wait for this licensing process?

Tim, go ahead.

Sorry, Morris. Yes, I can answer to that. So, in the near term, obviously, it has taken some extra time to get these permits. And the customers are working in every channel they can to get material in on time. But we do continue to receive permits. And if we continue to receive permits for -- specifically for those larger orders, we have a very healthy backlog that's ready to ship. And we believe that the market is just growing too fast to be adequately serviced by just 2 players at this moment. So long term, I think the business still holds good.

Yes. If I may add to that point, I think indium phosphide is at a critical juncture at this point, I think. I think indium phosphide traditionally has been serving the faster data center activities such as transceivers. But now with AI going from 800G to 1.6T to 3.2T, as the speed goes up higher and higher, the need for indium phosphide is more and also the requirement for lower EPD material becomes that much more important.

And so not only I believe with AI's advancement in data center applications will increase the need for indium phosphide tremendously, and also because of the device size becomes larger and the power requirement for these higher speed indium phosphide devices will need better quality material. And that -- all that said, should increase our indium phosphide business opportunity for AXT.

And with all that said, AXT is also a significant player on indium phosphide substrate supply overall. We believe we are either #1 or #2 in the world of indium phosphide substrate supply, which we estimate to be at least 40% of the world indium phosphide supply. So, I think although the permitting and the geopolitical restriction is hurting our business at this point, but I think the demand is there, we believe we should recover.

Great. That was really helpful. And then I have another question. It looks like it's great to see your gross margin improving again. And I wanted just to confirm, I remember in the first quarter, there had been a bit of a yield issue associated with germanium arsenide for a wireless HPT customer. It seems like that's been resolved. And have you resumed the business opportunity with that customer at this time?

Yes, we have. But although, we're taking -- as we said in the script, we're taking a fairly conservative approach. And so, we're not taking a big portion of it, but we want to not only serve the customer well, also holding our -- improving our margin.

So, I think we should be able to continue to see the improvement throughout the second half of the year on that business. And if we can improve the margin, so that also implies better yield and efficiency, we should ask for a higher portion of our business with that customer. So that should grow our revenue as well.

Next, we'll move to Tim Savageaux at Northland Capital Markets.

I want to go back to something Tim said or the other Tim, about the market growing too fast to be serviced by 2 players. And I want to kind of dig into that a little bit more. Obviously, we've got a lot of indications of that, both from what the hyperscale guys are planning to spend and what we're hearing from various members of the technology kind of ecosystem, demand seems to be pretty strong.

I wonder from AXTI's perspective, any more details on what you're seeing there in terms of the growth and -- or growth potential on, and whether that's how your backlog may have increased during the quarter, given the export issues and -- and how you see that playing out for the rest of the year? I guess, were we not facing these export issues in indium phosphide, is the growth rate that you had been looking for before -- has that accelerated? What are the trends there?

Thank you, Tim. It's a good question. So yes, I just basically want to repeat a little bit what Morris has just said. Obviously, we are seeing market trends that the demand for optical interconnectivity is growing rapidly. The move to higher-speed transceivers is moving rapidly just as we expect. This not only -- we've said for a long time now, this has a bit of a double benefit for us, because we are not only seeing that people as we move to larger and larger -- higher and higher speeds, the constraints on the lasers and detectors become higher and higher.

So higher quality material is required, lower EPD material is required. So, we also see some market share coming our way because of that. But in addition to that, a lot of these new devices are large. As we move into some of the larger EML devices and we move to silicon photonics, the acreage of indium phosphide that is used goes up, too. So, we do see the demand for indium phosphide substrates increasing very healthily, certainly at least at the growth rates that we were predicting earlier this year and probably even higher.

As Morris again said, we own about 40% market share in this. So, when I say, it's difficult for this market to be served by 2 players. We've got a quality and technology improvement over our competition. We've got 40% market share already. It's very difficult to fill that hole very quickly. We are still seeing orders, although the permit process is going slow. We're receiving new orders on pretty much a daily basis for indium phosphide. So we're definitely seeing the demand for AXT is still there. Once we start getting these orders -- these permits come through, I'm sure we're going to see more demand coming our way.

Okay. And then just a follow-up. I think you mentioned the $10 million backlog for both indium phosphide and gallium arsenide. And I guess looking at the shortfall in the quarter, I think you said half of that was gallium arsenide export. So, should we infer from that, that the majority, not the vast majority of that backlog is indium phosphide? And without the permits, I mean, could you ship all that this quarter? Or I guess, how quickly do you think you can get back to $10 million or get to $10 million a quarter in indium phosphide substrate revenue?

Right. Again, good question. Thank you, Tim. So yes, more than 50% of backlog is indium phosphide. And we've got a large amount of capacity there right now. We're typically turning orders around in 4 to 6 weeks. Sometimes, if we need to, we can turn them around faster than that. Of course, before all of this permit procedure came into place, we've got a lot of WIP, and we've built up WIP waiting to go to get some of those permits coming in as well.

So, it is possible that we can turn all of this backlog around pretty quickly. But of course, it's going to be very dependent on the timing that those permits come in and how they come in throughout Q3. But we do anticipate that this will -- if we get more permits, we're confident we will, we'll see an upside to our Q3 guidance.

We'll go next to Richard Shannon at Craig-Hallum Capital Group.

This is Tyler Anderson on for Richard Shannon. Could you expand upon why the gallium arsenide export licenses slowed down as the indium phosphide licenses began to be issued? And is this the same agency that's issuing these? And are you expecting any lower cadence of the indium phosphide licenses than what you expected before because of the gallium arsenide slowdown?

Again, I can answer -- I can at least start to answer and then Morris can probably elaborate a little bit more. It's difficult for us to speculate really what is going on here. What we do know is that it is not AXT specific. The whole industry has been faced with delays with gallium arsenide permits. What we have seen, however, is that through Q3 and certainly through this past month, the permits seem to be -- the permit approval process seems to be speeding up again.

And we've received quite a few more permits just in July. But I think it looks like they're now catching up through some of the backlog that we've seen there. So hopefully, things will return to normal fairly soon, both on gallium arsenide. And then hopefully, we'll see the same kind of cadence on indium phosphide as we approach normality on gallium arsenide.

Yes. I think Tim doesn't want to speculate, but I can sort of tell you, if you see the news that the ongoing of the restriction of rare earth in China implementing the policy probably has something to do with it. In other words, China will want to use this as a negotiating tool. So, I think they started to restrict the number of permits. But I think we are seeing the latter part of it start to relax more now. I think it's -- hopefully, it's getting into a more regular session that we should be able to get more permits regularly.

And are you seeing any sort of advanced order makings where customers are starting to build inventory? And could we see any kind of spikes in your revenue moving forward as you work through the backlog and people start to place larger orders while they can get a permit?

I would tend to think there are they are threatening to give us big inventory orders to anticipate the -- getting the permits. But I don't think we are at that stage because we're not delivering even the first big large orders. We do have a lot of very urgent order needs to be delivered. And so the customer said, okay, if we could get the first order through, they will give us other anticipated order, they want to build inventory. But I don't think we are at that stage to worry about that yet, because we're not even delivering the first batch. I mean, as far as big orders is concerned, we've so far delivered, we said $700,000 worth of indium phosphide orders outside of China. But the backlog is 6x, 7x or even 10x that.

And that concludes our Q&A session. I will now turn the conference back over to Leslie Green for closing remarks.

Thank you all for participating in our conference call. We will be participating in the Annual Needham Virtual Semiconductor & SemiCap Conference in August and hope to see many of you there.

And as always, please feel free to contact us if you would like to set up a call. We look forward to speaking with you in the near future. Thanks.

And this concludes today's conference call. Thank you for your participation. You may now disconnect.