Aehr Test Systems

Thank you, operator. Good afternoon, and welcome to Airtest Systems' first quarter fiscal 2025 financial results conference call. With me on today's call are Airtest Systems President and Chief Executive Officer, Gain Erickson, and CFO, Chris Yu. Before I turn the call over to Gain and Chris, I'd like to cover a few quick items. This afternoon after market close, Airtest issued a press release announcing its first quarter fiscal 2025 results. That release is available on the company's website at AIR.com. This call is also being broadcast live over the internet for all interested parties and the webcast will be archived on the investor relations page of AIR's website. I'd like to remind everyone that on today's call, management will be making forward-looking statements today that are based on current information and estimates and are subject to a number of risks and uncertainties that could cause actual results to differ materially from those in the forward-looking statements. These factors that may cause results to differ materially from those in the forward-looking statements are discussed in the company's most recent periodic and current reports filed with the SEC. These forward-looking statements, including guidance provided during today's call, are only valid as of this date, and their test systems undertakes no obligation to update the forward-looking statements. And now I'd like to turn the conference call over to Gane Erickson, President and CEO.

Thanks, Jim. Good afternoon, everyone, and welcome to our first quarter fiscal 2025 earnings conference call. Thanks for joining us today. Just as a heads up, we do these live, and I managed to have a cough with a little bit of a tickle, so I don't think I'm going to be able to get through all of this without coughing. But if I'm good, I'll hit the mute button and you won't catch them. But I apologize if I cough in the middle here. I'd like to start with a quick summary of the highlights of the quarter and spend some time giving an update on what we're seeing across the key markets areas addressing for semiconductor test and burn-in, including increasing traction in some new markets. I also plan to spend a little time on our latest expansion in the burn-in market with our acquisition of NCAL Technology this last quarter. Then Chris will go over the financials in more detail, and after that we'll open up the lines to take your questions. We finished the first quarter with revenue and non-GAAP net income ahead of consensus estimates and are off to a good start to our fiscal year. Silicon carbide wafer-level burn-in test systems and full wafer contactors are poised to be key contributors to revenue against this year. But we're also forecasting material bookings and revenue contributions from several other markets this fiscal year, as we're successfully executing our strategy to expand our test and burn-in products into other large and fast-growing markets such as artificial intelligence processors, gallium nitride power semiconductor, hard disk drive components, and flash memory devices. My plan is to cover each of these markets, beginning with the wafer-level test and burn-in for silicon carbide devices. We've been seeing a stabilization and increasingly positive discussions within the silicon carbide power semiconductor market over the past quarter. The silicon carbide market continues to feel like a typical new technology adoption cycle, which often starts with early enthusiasm and rapid growth projections from early adopters, followed by a period of adjustment and slower growth, and a bottoming of that growth before higher growth resumes to new highs. Electric vehicle suppliers are clearly moving towards silicon carbide in integrated modules, combining silicon carbide MOSFETs into single packages to meet the industry's power, efficiency, and cost-effectiveness demands. Some EV suppliers are shifting to integrated modules containing up to 32 or more devices in a three-phase inverter module. Due to the need for extensive test and burn-in of these devices to ensure reliability for mission-critical applications such as EVs, The benefits of conducting this screening at the wafer level before integrating them into modules are very clear. The process improves yields, reduces costs, and this is driving demand for wafer-level burn-in, an area where Airtest stands as the low-cost leader and proven solution for this critical testing. In recent weeks, AIR has presented at several industry shows, including the Power Up Conference in Italy three weeks ago, as well as the ISCS power semiconductor and the ice cream conferences in the U.S. over the past two weeks. The overall message from these conferences is that EV manufacturers are standardizing on modules, with some even moving towards entire integrated inverter systems supplied by the semiconductor companies themselves instead of the traditional tier one suppliers of the past who served as a middleman. The key takeaway from these trade conferences, as well as conversations with EV auto OEMs, is the critical importance of the testing and burn-in process to eliminate early life failures that would otherwise occur once these devices are integrated into inverters or full drive units in electric vehicles. Silicon carbide suppliers have varying opinions and standards when it comes to quality, leading to significant shifts in market share as some suppliers struggle to meet the integration and quality demands of EV manufacturers. For AIR, this presents a positive opportunity as the market shifts towards suppliers that are capable of delivering integrated systems and high-quality products. We're making steady progress on our previously announced benchmarks and engagements with new silicon carbide device and module suppliers. We're confident that we will add several new silicon carbide customers this year, establishing our solution as our tool of record for volume production. Additionally, silicon carbide is gaining traction in applications beyond electric vehicles, such as solar, industrial, and data centers, which will expand our addressable markets. We're highly optimistic about our silicon carbide business and expect it to gain momentum over the next few quarters. Our silicon carbide customers are forecasting capacity expansion needs in calendar 2025 with several anticipating purchases of one or two systems in early 2025, followed by production volumes in the second half of the year and ramping further into 2026. Separate from incremental system capacity increases, We continue to see strong demand for our FOX wafer-packed full wafer contactors for silicon carbide and other markets for our installed base of FOX-XP and FOX-NP wafer-level test and burn-in systems, driven by a record number of new device designs started this past quarter. These designs are expected to lead to additional wafer-pack purchases for engineering qualification, as well as to volume production orders as they advance to production. We had another solid quarter for wafer pack sales, generating over $12 million in revenue from wafer packs alone in the first quarter. Now let me move on and discuss our progress with testing and burning in gallium nitride power semiconductors. We're now in negotiations with our first gallium nitride customer, or GAN customer, for volume production wafer-level test and burning of their devices. We've been working closely with them and over the past year they have purchased and we have delivered a significant number of wafer packs to successfully qualify a wide range of GAN device types aimed at multiple different markets, including consumer, industrial and automotive. In addition, we had increased discussions and engagements with multiple potential new GAN suppliers about their needs. The data collected from multiple GaN companies shows that these devices will need burn-in to meet the requirements of many of the markets they're targeted for, particularly industrial and automotive applications. We believe GaN is a significant up-and-coming technology for power semiconductors. With a forecasted compound annual growth rate of more than 40% to over $2 billion in GaN devices sold annually by 2029, GaN has the potential to be a significant market opportunity for AIR's wafer-level solutions. It's also interesting to note that several of our current and prospective customers value the fact that AIR can provide proven test and burning capabilities that successfully address both silicon carbide and the GaN devices. The testing of each is quite different. However, we can address the requirements of each market with the unique proprietary features and capabilities in the same system on our Fox platform. We continue to make investments in our platform with new capabilities and enhancements that enable us to provide value to these customers, expand our markets, and maintain our leadership position as a low-cost proven solution. So moving on to the artificial intelligence processor market. Last quarter, we announced that an AI accelerator company committed to evaluating our Fox solution for wafer-level test and burn in other high-power processors. This evaluation is very far along at our Fremont facility, where multiple wafers are being tested using our proprietary wafer packs and new high-power Fox XP and NP systems, which provide up to 3,500 watts of power delivery and thermal control per wafer. We are delivering over 2,000 amperes of current to a single 300 millimeter wafer, allowing us to burn in a large number of processors in parallel with our proprietary test modes. This company is working closely with us to be able to use the Fox XP system for wafer-level production burn-in of their processors, which would prove to be more cost-effective and significantly more scalable than doing the screening later in their manufacturing process. We think this presents a significant opportunity to displace the current package and system level tests for AI processes for large language model development. And we believe we can meet this enormous challenge with the current capabilities of our new high power Fox XP system. This evaluation is progressing very well. And once we demonstrate successful wafer level test results and throughput, we anticipate they will adopt our high power Fox XP systems for production of the next generation AI processors beginning this fiscal year. As we've noted before, based on the production forecast, we believe they could potentially be more than a 10% customer for us this fiscal year alone. Turning to our acquisition of in-cal. During the quarter we announced and completed our acquisition of in-cal technology expanding our product portfolio to include In-Cal's highly acclaimed package part reliability burn-in and test solutions, particularly its ultra-high power capabilities for AI processors, GPUs, and high-performance computing processors. These advanced high-power capabilities, combined with AIR's industry-leading lineup of wafer-level test and reliability solutions, position us strongly to capitalize on the significant and rapidly growing opportunities in the AI semiconductor market. This acquisition greatly expands our addressable market, enabling us to provide a comprehensive turnkey solution for reliability and testing from engineering to high volume production to the rapidly growing AI semiconductor market. We're excited to bring the combined strengths of both companies to market as we begin engaging with Intel's customers, including many AI industry leaders. Customer feedback on this acquisition has been overwhelmingly positive, with several meetings held over the past few weeks where some customers indicated increased forecasts for engineering qualification as well as for volume production. Last month, we were pleased to announce the first volume production orders for Intel's new Sonoma UltraPi power semiconductor test and burn-in solution for test of AI processors. These orders were placed by a large-scale data center hyperscaler, that provides computing power and storage capacity to millions of users worldwide today. We've also had many customers express interest in In-Cal's medium power Tahoe system, which is being used today across multiple market applications for both qualification and volume production. This includes multiple medical companies that use the Tahoe system's special capabilities for production test and burn-in of devices in a critical medical device application. This speaks to the special capabilities of these systems and the quality of their hardware and software to meet the critical traceability requirements of the U.S. Food and Drug Administration. The integration with NCAL is progressing well, and employees from both companies have been very positive about the combination. We have already shipped several systems since the acquisition, and with NCAL being just down the road from us, We plan to consolidate personnel and manufacturing into AIR's much larger Fremont facility by the end of the fiscal year. As we've mentioned before, we're underway with an upgrade and remodeling of our Fremont headquarters this year, in fact right now, and are ensuring our facility infrastructure meets the needs of both companies. So turning to the hard disk drive market. Last quarter, we also announced a key customer in the hard disk drive space that is now forecasting a production ramp up starting this fiscal year for a new high volume data storage device application. This customer is finalizing their capacity requirements and we expect this ramp to drive orders for multiple Fox CP production systems and wafer pack contactors with shipments likely occurring in the second half of this fiscal year. As we've noted before, they could even be a 10% customer for us as well. As some of you may recall, this is a major customer that we announced in 2019 pre-COVID that purchased our Fox CP, our single wafer test and burn-in solution, for wafer level test and burn-in of their devices in this very high volume application for enterprise and data center market. We see the data storage market, along with various devices supporting the global 5G expansion, as new growth opportunities for our systems, as these markets require devices with exceptionally high levels of quality and long-term reliability. And lastly, let me talk about the NAND flash memory market. We're excited about the ongoing progress of our benchmark with a leading flash memory manufacturer we announced last quarter. This benchmark is aimed at assessing the potential of our Fox XP solution for wafer-level test and burn-in of their flash memory devices. This evaluation focuses on providing 100% test and burn-in for devices intended for high reliability applications, such as enterprise storage. We're currently developing a new high pin count fine pitch wafer pack that can support flash memory requirements, but can also support DRAM testing should customers choose to pursue DRAM burn-in in the future. Our objective remains to complete the proof of concept by the end of this fiscal year, positioning us to secure a commitment from this customer to develop a production test cell with the potential to contribute to revenue in our fiscal 2026, which begins next June. As I've noted before, we believe this is the front end of an exciting and potentially enormous opportunity for our solutions, and we see the NAND flash market as a key market opportunity with long-term potential to also move into DRAM wafer-level test and burn-in. With all these customer engagements, market opportunities, and the products to address them, we're very optimistic about the year ahead, and we're reaffirming our financial guidance for revenue growth and profitability for the year. With that, let me turn it over to Chris before we open up the line for questions.

Thank you, Gay. Before I turn to a review of our financial results, I'd like to know that over the past 12 months, Bayer has invested financial and human resources in improving our own infrastructure to support the continuing growth of our business. We recently went live with a new enterprise resource planning system in September. That will provide more sophisticated functionalities and capabilities to support decision making and compliance. We expand our policies and procedures in various functions to provide more internal control and structure. We have made improvements in our supply chain and engaged with contract manufacturers that offer more value at the quality components and services to support our growth. I would like to thank our teams for their dedication and strong execution. With a more robust infrastructure in place, we're more ready than ever to support both our organic and inorganic growth. Also, as Skye noted, the integration of our in-cal acquisition, which closed on July 31st, is progressing well. Our plan is to consolidate personnel and manufacturing into AIR's Fremont facility by the end of the fiscal year. With the upgrade and the remodeling of our Fremont headquarters underway, we are ensuring our facility infrastructure meets the needs of both companies. Now turning to our Q1 performance, which included one month of financial results from the in-cal acquisition. Our Q1 results exceeded analyst consensus on both the top and bottom lines. Under the current challenging market and macroeconomic environment, first quarter revenue was 13.1 million, down 36% from 20.6 million in Q1 last year. Demand for waiver packs contributed to the majority of our total revenue in the first quarter. WaverPak revenues came in at $12.1 million and accounted for 92% of our total revenue in the first quarter, significantly higher than 55% of the prior year's first quarter revenue. WaverPak revenues continue to represent a significant recurring revenue stream for our business, as our customers continue to utilize the available installed base of Fox XP systems for new customer design wins. and purchased additional new wafer packs from AIR to test and burn in these new devices. Non-GAAP growth margin for the first quarter came in at 54.7% compared to 48.7% year-over-year. The increase in non-GAAP growth margin was primarily due to favorable product mix of higher margin wafer packs. Non-GAAP operating expenses in the first quarter were $5.5 million, essentially flat compared to $5.4 million in Q1 last year. We incurred income tax expense in Q1, and our effective tax rate was 18.9%. Non-GAAP net income for the first quarter, which excludes the impact of stock-based compensation, acquisition-related costs and amortization of intangible assets acquired through the in-cal acquisition, was $2.2 million, or $0.07, per diluted share for the first quarter. Compared to non-GAAP net income of $5.2 million, or $0.18 per diluted share, in the first quarter of fiscal 2024. Our backlog as of Q1 was 16.6 million. Now turning to our cash flows and the balance sheet. We generated 2.4 million in operating cash flows in Q1. Our cash, cash equivalents, and restricted cash were 40.8 million at the end of Q1, down from $49.3 million at the end of Q4 last year, resulting from the 10.6 million of our cash on hand used to fund the acquisition of in-cal technology during the quarter. We have zero debt and continue to invest our excess cash in money market funds. Interest income earned during this high interest rate environment was $681,000 in the first quarter, up 17% from the same quarter last year. Our previous S3 registration statement, which was filed in September 2021, expired recently. Since we did not sell any shares through the ATM offering during the open trading window in August, about 42.7 million of ATM that could have been sold under the previous S3 was not utilized. As part of our good corporate governance practices, we plan to file the new S3 with the Securities and Exchange Commission soon to support potential future financing needs. Once approved, the new S3 self-filing will be good for three years. Now turning to our outlook for the current fiscal 2025 that ends on May 30th, 2025. As Gay mentioned, we are reaffirming our previously provided guidance for the fiscal year of total revenue of at least 70 million and net profit before taxes of at least 10% of revenue. We expect approximately 85% of our forecasted annual revenue to come from waiver level burning and around 15% from package part burn-in driven by our acquisition of Incal. Lastly, looking at the investor relations calendar, Airtest will meet with investors at the LD Micro main event in Los Angeles on Tuesday, October 29th. And then the following month, we will participate in the Craig Hallam Alpha Conference in New York on Tuesday, November 19th. Then on Tuesday, December 17th, we're heading back to New York City to attend the NYC Summit. After the new year, we'll participate virtually at the Neom Growth Conference on Thursday, January 16, 2025. We hope to see some of you at these conferences. This concludes our prepared remarks. We're now ready to take your questions. Operator, please go ahead.

Thank you. At this time, we will be conducting a question and answer session. If you would like to ask a question, please press star 1 on your telephone keypad. A confirmation tone will indicate your line is in the question queue. You may press star 2 if you would like to remove your question from the queue. For participants using speaker equipment, it may be necessary to pick up your handset before pressing the star keys. One moment, please, while we poll for questions. Once again, please press star 1 if you have a question or a comment. Okay, our first question comes from Christian Schwab with Craig Hallam. Please proceed.

Hey, this is Christian. Hey, Gabe. So can you, on the silicon carbide, the additional customer commentary about getting two additional customers this year, are these customers that you've been working with in the past? And alongside that, can you give us an idea of how you would anticipate that ramping in near term and then over time? I mean, could they become as material as your current largest customer in silicon carbide?

Okay, sure. Yeah, it's actually some of the same old story of some of the folks that we anticipated would have bought their first systems last spring, really. that we're now at basically the ramps and decision timelines have pushed. So I'd say two or three of those are actually sort of the same guys. That's just, you know, kind of a year later. But there's also two, I think in our forecast anyhow, or high likelihood forecast that are, were not anticipated last year that have, you know, you know, come online over the last, we've been working with them over the last, say, three to six months, engaged in paper benchmarks and other financial things that are also looking very good. So I wouldn't ask you to go say I'm going to win six new customers, but there's the potential for it. The interesting thing is what it feels like from you know, probably their end markets, why they all feel similar is everyone seems to be talking about putting, you know, like first tools in place somewhere between the end, literally the very end of this year till, I don't know, spring timeframe till maybe May, kind of within our fiscal year, which we think that wind feels like it's, you know, maybe one or two systems. and then they ramp into the fall based upon a number of the EV data center, even maybe some potential solar winds that are driving capacity by that timeframe. And that aligns with the physical buildings. One of the things that was going on about a year ago, there's sort of this excitement, I'll use that word, around how fast everything's going to be adopted. I'm looking at a you know, a land of dirt, and it's like that building is not going to ship MOSFETs next year. I don't know what people are thinking, but it's too aggressive. Now, you know, the forecasts seem to align with the physical buildings, with the OEMs that are, you know, describing which cars they're going into, et cetera, you know, and are still forecasting those cars, you know, moving forward. And this includes, you know, OEMs that include China, other Asia, Korea, U.S. and particularly European suppliers. So I hope that helped to explain it.

I guess it still wasn't clear to me.

Oh, just five. five of your five potential or six potential customers um are you know can you give us like real color yeah the size you're right i forgot to say that i i didn't actually didn't mean to drop that so um i i would say you know one of the things that and you you heard it in my prepared remarks It's very, you know, and I have been constant at these public conferences. Vernon has been asked to speak as well. We have tested enough wafers. I mean, we've tested more silicon carbide customers' wafers than anyone by far to where we have a huge cross-section of data, and we just simply make the statement that if you want to do automotive quality down to sub-part per million failure rates of these things, You need to do an extended burn-in, not just a simple screen that's measured in hours. And I would say that each of the customers that we're talking to right now, that's what they're saying. They're all talking about, you know, these extended multi-hour 6-12 plus even up to 24 to begin with hour burn-in times. That'd be consistent with the automotive guys and it's being repeated by the OEMs as a critical requirement. So with that, you can do the same modeling that we did before, which is okay. It's a way for fab. It has 4000 way for starts per week, let's say, or whatever it is. And you multiply, you know, you say it's got a six or 12 or 18 hour burden time. and 18 wafers. So these are customers that have the potential to buy significant volumes, perhaps even as big as our biggest customer to begin with. We also believe that our largest customer is doing extremely well in the market and is going to grow substantially both in capacity and design wins over the next several years. So they seem to be doing very, very well in the market.

Okay. I guess this is my second last question. On the AI processor opportunity, I know last quarter you highlighted, of course, it is not NVIDIA that you're dealing with. But can you give us any type of color there, you know, given, you know, a 10% customer value? It's fantastic, but that's only like $7 million worth of sales. I'm kind of thinking that that should be a much more substantial number than that over time. But maybe you can give us a little bit of color of the open-ended opportunity there.

Yeah, so, you know, we can track the units. We understand the test time. The test times on processors are quite long. We know it's 100% burn-in, and we know and can track failures and see what's going on, and there's some interesting tricks and things that are going on there. So the data is all supporting the capability and feasibility of doing wafer-level burn-in and you know, we can see, and now, candidly, through the in-cal acquisition where we're testing a large number of AI processors, you can see what they're testing and why they're testing it. No one's getting out of burn-in of AI processors anytime soon. So, you know, we haven't given any more color related to which of the processor, guys, it is. We have said, you know, we just simply said it's not NVIDIA because it just would be unfair to throw that around and have people imply that. But it is a revenue generating AI processor company today. Okay. And, you know, it's certainly, it's currently different than any of NCALS customers. That doesn't help you because I haven't told you who they are either. But we've actually had, in one of the conversations we had with one of the Intel customers that are currently testing AI processors on their systems, they leaned forward and were super interested to hear more about our wafer-level burn-in, which is not a surprise to us. And so, you know, I think that there'll be a mix, and we're not Picking sides, there's some real advantages of moving the AI processor and ultimately the DRAM to wafer level. But there's also some real challenges with it, depending on customers and test modes, et cetera. And so we can serve them with the package part burning capabilities of NCAL. But, you know, everybody, I mean, most people have heard the story, go back and listen to it. You know, our whole strategy has been that with this massive increase in semiconductors that are, in fact, getting less reliable, showing up in more applications where reliability matters, and then people putting them on multi-chip modules, it's a huge opportunity for test and burn-in, in particular for wafer level. I noticed that today, just probably, you know, AMD introduced the, what, MI325X. Did you see it? It looks like it has eight core processors and another eight stacks of HBM memories. You know, what happens with one of those core processors dies and burn it? So the value proposition concept of moving it from package to wafer totally still makes sense. But now we also have both. So little pitch for us as a company. And I can tell you, I know I use that phrase if you've heard me, I usually like to say we have a front row seat. Guys, we have the front row seat. This is a fantastic, I have sat in meetings in the last four to six weeks with some of the names of the names and hearing about their requirements and their actual process that we're testing, the next gen, the one after that. We're engaged with them to deliver their requirements. They're very excited because of our capabilities in thermal, our engineering capabilities, our manufacturing capacity capability, and our application support around the world. So I'm very excited about it, and I think there'll be leverage in both directions. I know we can already use some stuff from air to make NCAL's products better, and I think there'll be opportunities for customers that are using NCAL that may want to do a wafer-level burn-in step as well. from us. So I couldn't be more excited about this and it's super fun.

Thanks, Chris. You bet. No other questions.

Thank you. Okay. Thanks. Operator?

We caught him sleeping. Hello?

Oops, sorry about that. I had a malfunction with the mute button. Once again, if you have a question or a comment, please indicate so by pressing star 1 on your touchtone phone. The next question comes from Jed Dorsheimer with William Blair. Please proceed.

Hey, yeah, thanks for taking the question, Gain and Chris. I guess first question, Gain – Thanks for the clarity in Chris. Thanks for the breakdown for the year on the air business. Um, you know, I haven't seen, uh, previously you would press release, um, uh, wins and, you know, I know that you did on the wafer pack that kind of flowed through this quarter, but I was wondering in the 16 million of backlog, could you give any further detail in terms of the contribution of mixed there? is we haven't seen any press releases that's, you know, to get an idea of what that constitutes.

Yeah. I'm going to take a pass on that right now. I get it, Jed. I understand. We haven't, in fact, for those, someone had to have noticed, I think this is the first quarter in eight quarters, we didn't do the effective backlog. I don't know if everybody else noticed. We're really struggling with that. We've, we kind of, cornered ourselves here in that the only press release we did this quarter was the production Sonoma production system, right? And we quantified it. We said it was six systems. Then, as we went to look at the backlog and the breakdown, it was like, oh, it highlighted too much. You can imply that most of the incremental sales in the month were from that. And so we just chickened out and said, we're not going to say it because we simply don't want to triangulate and give precise details. Data to our competitors to their competitors, et cetera, because actually in the industry, people sort of figured out who it was. So I apologize for that. But, you know, specifically in the in the air business, if you will, you know, we have been material orders. We have been press releasing. The materiality is kind of increasing as we've been getting more and more wafer pack orders coming in that don't seem to press release out of that. We're still trying to balance that. I would consider any new production win as we are forecasting material and we will announce it. Again, we'll either typically give you a price and be vague about the quantity or give you the quantity and be vague about the price, just sort of as a process rule. But I will share that actually, you know, a reasonable chunk of that backlog is from NCAL. And, you know, NCAL's forecast is looking, I'll call it in the bag, if and with, you know, a reasonable amount of upside. We'll see as we get closer. But they have much longer lead times than us. We'll probably tighten that up a little bit. But they typically have about a six-month lead time. And so, you know, everything for this quarter is in backlog, for example. In fact, some of the next quarter is already in backlog for those guys, whereas AIR tends to run much shorter lead times. And the result of that is you have a little less visibility. And I think they're probably – better off than us, but the customers do like our short lead times.

Got it. That's helpful. I guess maybe, and I know you, uh, you don't want to give more, but let me just ask it anyways. Um, if I look at the wafer pack, which was all of the, you know, uh, product revenue in the quarter, um, do you see a more balanced in this current quarter? Is it more balanced with respect to system leaving Intel aside system. And yes, yes, yes, for sure. For sure.

Yeah. And they shift towards, you know, non-sick systems too, which is pretty fun. right so not you know we'll start seeing uh we're forecasting um some of the the other areas um you know it's it's actually i mean we we really never were all you know only about silicon carbide it was just fantastic we still love it we still love that market but the other markets are heating up and and you know consuming much of every single day for us candidly got it and then

What's your attach rate on the automated – what do you expect the attach rate for automated aligners with the systems? Do you see that as – Yeah, that's it.

We were actually just discussing that today. It's still a mix. We have customers that are forecasting systems with manual aligners still. I personally love the automated aligner, but it costs a little bit more. And you can have an automated aligner that is offline to feed multiple chambers, or you can have an automated aligner that is integrated with the system. And then we have several customers that are forecasting fully integrated. Absolutely, positively, nobody can do anything. I don't want to touch it. You know, lights out operations. So it's a mix going forward. It's a mix going forward. And by the way, by contrast to two years ago, it was all manuals. And I'd say as we go forward, I feel like it should be more than half would be automated aligners, but there's still some customers that are talking about manual aligners.

Okay. And I'll just keep going with a couple more here if you don't mind. What's the mix for your silicon carbide visibility or expectations between 150 versus 200 millimeters?

Well, we have a lot of 200-millimeter wafers that we're testing today. We have 200-millimeter wafers we're testing today. Vernon might tell you we've tested – anyone who's made a 200-millimeter silicon carbide wafer has been tested in this facility and most likely at their facility on our equipment, if that helps you with some hints. Of course, we haven't told you who they are. You know, I think our data, you know, seems pretty consistent with what's out there. There are 200 millimeter wafers. They're coming out. It's still the dominant is the 150 millimeter wafers. We think that, you know, 200 will become more and more, particularly in 2025. And, you know, from our system, for people listening, the Fox system itself can test all size wafers. 4, 6, 8, and 12 inch, or if you're so inclined, 100, 150, 200, and 300 millimeter. The wafer pack can work with any of them. We can actually design, in the same system, you can be testing 150 millimeter silicon carbide and 200 millimeter carbide. You just have to design the wafer pack contactor for the bigger wafer, okay? or the smaller wafer. So it's sort of a don't care to us, and we wouldn't have as much visibility as, say, a tool supplier who only sells 200 or 150, and they make or break. The customers, it's probably not as dramatic for us. So it's coming. I would probably leave it to you guys to have better visibility of exactly when the cutover is.

Got it. And then last question for me on... You know, is you historically new customers have tended to purchase a Fox NP for evaluation before switching over to the XP. And we haven't seen many NPs, you know, leave the docks. So what are your expectations on some of the, you know, on the ramp of these new customers, both in the silicon carbide, as well as in, in, other applications in terms of how that cadence will look?

Just give me a second here. I'm not coughing. I think four of the next five customers start with XPs. Okay. All right. Now I coughed. You got me there. Yep. Yeah. So it's interesting. So they're kind of skipping it. in some cases. Now I could be wrong folks, but, and we like, I mean, they coexist, but yeah, I mean, I think Vernon's got some quotes out to people that just say, I'm just going right to an XP. So, yeah.

Got it. Sorry. Last one. I promise. Is that because you do the testing you've been, I know you have this program where you have that, uh, you know, dedicated room where you do some in-house testing for your customers. Do you think that's what's shortened, or do you think it's a greater comfort around the process? Okay.

Both, honestly, but both, for sure. I mean, we're testing their way. For some of those people, we've been testing it for God knows forever, right? We've done full characterization, process control. We've done, you know, qualifications, the whole thing. They kind of, they're like, all right, skip it. Let's go. We get it now, you know? So there's a level of comfort. Um, you know, and us testing the wafers has a lot to do with it.

Appreciate all the color games. Thanks.

Once again, if you have a question or a comment, please indicate so by pressing star one on your touchtone phone. The next question comes from Matthew Winthrop with Equitable. Please proceed.

Hi, Gann. How are you, sir?

I'm good, sir.

Congratulations on, uh, on all the excitement in your outlook. You know, I had a quick question. I was looking at last quarter's release, and you always make a point of saying a customer or a large customer of ours. You don't usually mention names. And there's this sentence I've thought about where you say optical IO or co-packaged optics for companies like NVIDIA, AMD, and Intel, which have all had discussions, blah, blah, blah. You've never used names like that or haven't at least in years. And I know you're not alluding that they're direct customer of yours, and you probably touched upon this with your meetings. But I was wondering, a quarter later, can you expand on that? And is that from the new acquisition of Intel? Thank you.

No, I mean, okay, so the context and what we've said in there for people that are listening in, is one of the areas, and we, in fact, on this conference call, this is probably the first time in many quarters I didn't have a subsection on silicon photonics. So silicon photonics is more than just putting a laser on a piece of silicon. It's actually the entire photonics engine for communication. And folks like IBMs and Intel's have literally worked on this for 20 to 30 years because silicon or electrical transmission line or electronics can only go so fast. There are physical limitations because atoms actually have mass to them. And we're within approximately a generation. We're running at, say, 200 gigabit transmission rates or 112. In this range, you're at a point where you actually cannot make the electronics go any faster. So people have anticipated that the bandwidth requirements of communication and data would get to a point where you are bandwidth limited. You literally couldn't make the transmission line faster or smaller or what. The answer has always been fiber optic transmission or photonics based. But the practicality and the cost of that had never really been coming together. Intel was one of the most visible companies who now, what, six years ago, finally figured out to sort of on a wafer scale, come out with a complete silicon photonics engine that would allow them to make the entire transceiver, if you will, in a chip. And that chip, they started by putting into fiber optic transceivers, little fiber optic clips and data centers. They successfully did that because of the semiconductor performance, if you will. And there's other companies that are trying to do or doing similar types of things, okay? Just in, you know, and I'm not representing Intel here, I'm just talking what's public. Intel has since sold the fiber optic business, the actual transceiver business, but kept the semiconductor business as what we believe from their public statements because of the criticality of it, of the need for it to go into co-packaged optics or to put it on processors and for chip-to-chip communication. When I started talking about this two years ago, it was barely even discussed. Intel hadn't even announced a product yet. But AMD and Nvidia were both alluding to it and their roadmap. And so I've been kind of repeating, you know, where is it going? Because we've made some investments and partnered with companies in anticipation of the potential for this widespread adoption of optical communication between chips. way different than just the data center. At that time, for clarity, and it's kind of fun because you look back on it, the whole AI thing, this is the pre-ChatGPT, it wasn't clear where it was coming from. We were like, well, maybe it's going to be in servers. It would certainly be in the data farms once. What we've seen in the news now is folks like Intel, AMD, and NVIDIA have talked about the massive data crunch challenge with now the AI processing capability of these large language models in these clusters. How are you going to get this all done? And so we're now seeing roadmaps that talk about, well, what you do is you put this little optical I.O. part in the chip, and the chip-to-chip communication is not through an electrical system. transmission line, it's through an optical transmission line. That's it. So that's the name dropping. I'm not trying to say who and who isn't our customer, but it's those three in particular are the ones, the big dogs that are going to drive the roadmap. And if you go talk to market forecasters, they go, there's no market for it, or there isn't a market. Well, no, there isn't a market until NVIDIA, AMD, or Intel make one, and then there's a market. And then it'll be massive. So we've just been trying to explain to people, you know, why did we make these investments? I mean, the customers bought the equipment from us. It's good margin for us. But will it take off and be a big market? You've got to ask those guys. I'm betting, I would bet, and we have chosen as a company to bet, that market will take off. When exactly? I don't know. But things sure seem to be moving faster today. at a pace faster than historically. So, I got my fingers crossed. But we are ready and ready to go for this. We have the tools and we're very excited about it. And that same 3500 watt system that we developed for the optical AI is being leveraged for the actual AI processor, which is pretty cool.

Okay? That's terrific. So, suffice it to say, this concept is not, when you announce your reaffirming guidance for the year, This is all. Maybe someday in the future that so this is a potential upside.

I definitely say that. No one's asked me. I'll cough it up. We actually have very little silicon photonics revenue this fiscal year. I think we have a bunch of wafer packs and a bunch of design-related things, but I think currently we have very little capacity. But we've had some requests, and we've asked for lead times and things like that. We could actually ship a number of systems or upgrades of systems still in the fiscal year, It's the upside or downside of all the inventory we bought, candidly, but we have the capacity of being able to build all these systems. You know, if a customer asks, could you sell me one in a month, the answer is yes. You know, there's pros and cons to that, right? That means they can wait a month to ask, right? But I feel really good about our position.

Go get them. Thank you, sir.

Thank you.

If you have a question or a comment, please press star one. Once again, please press star one if you have a question or a comment.

Any other takers?

Okay, looks like we have no further questions in queue. You're very welcome. I'd like to turn it back to management for closing remarks.

All right. Do you think we missed anything, Chris, that we didn't cover? Okay. So, folks, thank you very much. I really appreciate it, people joining in. And as always, if you're near, we always invite you to come by. It's a little hectic and messy around here right now as we're doing some basic construction and things like that. But we still have customers coming through. I think there are three or four this week. We have Another one tomorrow. So we'll invite you in if you happen to be in the Silicon Valley. We'd love to host you. If not, maybe we'll see you at one of the trade shows or the investor conferences, and we look forward to giving you an update on the quarter as things go at the next quarter event. Take care.

This concludes today's conference, and you may disconnect your lines at this time. Thank you for your participation.