We will now begin the Q&A session. CLSA Securities, Yoshida-san, please go ahead.
Yoshida from CLSA. Thank you for taking my question. I would like to ask you about the quarterly sales momentum trend. I know that you are seeing an adjustment phase in the second half of the fiscal year, but could you please give us the current color of the quarterly outlook for the rest of the quarters in this fiscal year? Also, I think Douglas talked about the market outlook for 2026 in the last quarter as an optimistic view. Do you have any clearer view for next year's growth pace at present? Thank you very much.
Okay, thank you, Yoshida-san, for your question. The first question was about the color of the remaining quarters for the year. Clearly, our first quarter was a banner quarter, and there is going to be some digestion. We see that the lowest quarter for the year will be in the third quarter, with the fourth quarter resuming some growth trajectory that will then continue into FY 2026. That is sort of the profile that you can think about for the sales on a quarterly basis. As far as 2026 goes, you know what we have been able to say is just more qualitatively that we are very optimistic about 2026.
With the additional wafer capacity and advanced packaging being added into the market, that is going to drive additional unit volume and test capacity needs. We have also seen that the hyperscale CapEx growth is remaining intact, and that is also a very encouraging signal for 2026. As the device complexities continue to become more sophisticated, that requires additional test content. All of those things continue to give us optimism about 2026. We are hopeful at some point in time that the consumer and automotive segments also return to some level of growth. For those reasons, without giving actual numbers, we do feel more and more confident about 2026.
Thank you. Just for the clarification, you mentioned about the capacity addition of the 70% capacity addition, but when would be the time to fulfill this new capacity in the future? Thank you.
Yeah, thank you for the follow-up. I would say, number one, we're able, with our current capacity, to meet all the current needs of our customers, which we've put a lot of effort into over the last several years in expanding. As I mentioned in the prepared remarks, we've tripled the capacity in the last several years. As to this next step function in capacity additions, we're targeting for the end of 2026 to add both of this for our SoC platform as well as our memory platforms.
Thank you. That's very helpful. Thank you very much.
Thank you for your question, Yoshida-san. Next, we'll take a question from Damien Tan, Macquarie Securities. Please go ahead.
Congrats on the quarter. I just want to follow up on the capacity increase. A baseline year then is the last fiscal year. Can I understand that you are targeting, for SoC tests, for instance, revenues somewhere in the ballpark of the mid-$700 billion? Is that the capacity amount we will be looking at? On top of this, a second part of the question I'd like to ask is maybe you can give a hint on where you see the growth in the adjacencies that you talked about earlier, for instance, I presume including in services as well as, for instance, in, how do you call it, in design support activities. Thank you.
Thanks for the questions, Damien. Let me make sure I understand the two questions. One was just on the level of expansion. For SoC testers with our production capacity expansion. The second one, I believe, was about the engineering and test services business that we're developing. I'll answer the first one, and then I'm going to ask Sanjeev Mohan to answer the second one. On the first one, on the SoC side, the range of expansion is going to be in the 60%-70% range based on our 2025 levels. We're having a significant expansion of the capacity. One thing to keep in mind is we're just not building to the capacity that's necessary to only meet the next wave of demand. Obviously, we're going to put additional capacity in place in order to handle some of the volatility that we've seen in the past couple of years with unforecasted business.
We've learned that it's very difficult for our customers to forecast some of the spikes in these AI-related waves. For us to be able to respond quickly to those demands, we have to build in some additional buffer levels of capacity in order to meet those spikes. That's the basic number, about 60%-70% on the SoC from 2025 to end of 2026, where we see that being put in place. I'm not sure I caught, so we may have to clarify a little bit more on the second question, but I do want to introduce Sanjeev Mohan. Sanjeev is taking over for Mac Nakahara, who for many years led our global sales and support activities in the company. Mac is continuing with the company as a close advisor to me in the CEO office.
Sanjeev now, who's been overseeing all of the U.S., North America, as well as European sales, is taking over the global sales role for the company. I'm going to have Sanjeev answer the second part of the question.
As far as the adjacent growth, I mean, there are a few areas that we have been investing in. The SLT system-level test, we expect that business to grow in the future. As the nodes get smaller, the demand for system-level tests will continue to increase. As a company, we continue to invest heavily in that area. That is going to be one significant business that we expect to grow. We also have invested in our DI device interface business. We expect that business to continue to do well as more and more customers are asking for turnkey applications. Then there is engineering services or warranty support. As you can tell by our results, we are shipping more and more systems, and that helps provide organic growth for warranty support. That is another area that we expect in the future to do well and continue to grow.
In regards to other services, as a company, we have understood that our customers want more from us, and I think that provides opportunity for us to get into some services areas that we have not done in the future. I do not want to get into details of that because it is something that we are developing, but that is another area that we hope will add to our business results.
Oh, thank you very much. Can I just make one follow-up, small one? For the 60%-70% increase, can I just clarify the baseline for that? Is that average for 2025, or should we think of it from the end of 2025? Sorry, from the end of 2024, early.
Yeah, I mean, from a baseline level, we're. I would say, without quoting an exact number. For our 93Ks, we can ship more than 3,000 systems in a year. You can kind of use that as a reference point then to build off of. If that's helpful.
Thank you, Damien. Next question will come from Wadaki- san, Morgan Stanley MUFG Securities. Please go ahead.
This is Wadaki speaking. My question is, I guess, that the focus really is the earnings, the sustainability earnings, whether it is special demand or temporary demand. I think what I want to know is that, was there an element of special demand in first quarter sales? I also want to know if there is upside to the second-half forecast. For the first quarter, SoC testers, like if you break it down between a GPU and ASIC, can you give colors? That would help me kind of explore how sustainable the current momentum is.
Thank you, Wadaki- san. For the first part of your question about sustainability of some of the large numbers we saw in the first quarter, on the sales side, to a large degree, there were many pull-ins. We had originally not expected to have such a large quarter. We thought it would be spread out a little bit more, so there was a large degree of pull-ins. At the same time, the demand also was a bit higher than we had originally expected. That is kind of the top line theme there. On a profitability side, the stars sort of aligned perfectly for us in the first quarter. There was a really good product mix. There was a good degree of upgrades and licensing that always drive higher profits. As I mentioned in the prepared remarks, there were also some economies of scale and efficiencies.
We did not have any one-time write-downs. That all led to this really, really great result in Q1. Now, some of those things are sustainable, and some of them are not. The things that are sustainable are things like, as we look into the future, probably the product mix may not be quite as good, but still going to be very rich with the SoC versus memory. We feel like some of the efficiencies that we are driving are going to continue. Certainly, we are not going to expect the lumpiness over time to be perpetual across all of these different quarters. I think as we look into Q3 and Q4, some of the effect of those pull-ins are felt in the lower sales. Also, there is a lot of device transitions that are happening right now.
As some of the big AI accelerator, both on merchant GPU as well as the custom ASICs, are coming up, as well as the memory side with HBM transitions, there is a lot of work being done, investment being done to bring up those new types of devices that will show up later, more likely in FY 2026. As far as the breakdown between kind of traditional GPU-based accelerators and the custom ASICs, we are not able to really give you that split because it gets into customer-specific things that we like to stay away from. I mean, as you can guess, it is still largely on the traditional GPU basis, and we see the custom ASIC stuff growing now with a much larger proportion coming in 2026.
Thank you. That was very helpful.
Wadaki-sa n, thank you for your question. Next, Goldman Sachs Nakamura-san, please go ahead.
Yeah, thank you very much for taking my question. I just wanted to also touch upon the second-half guidance or maybe your September quarter guide as well as the second-half guidance. I understand that there was a lot of pulling into June quarter, but nonetheless, it seems like your guidance implies a significant decline into September quarter as well as second half. Is it a reflection of already a fixed sort of negotiation with your customers? Is it mostly a done deal to think about, or is there some conservatism also baked into their guidance given the lead time? That'll be my first question.
Yeah. So this is Sanjeev Mohan, Nakamura-san. As far as, so long-term, I mean, if we just step back for a minute, the business, the AI boom will continue. This goes back to the sustainability question. We think 2026 will be strong, and the business will continue to be very good. Q1 just happened to be much stronger than we anticipated. From our perspective, it isn't that there is a significant decline in the second half. It's more that Q1 was much stronger, and then there were some pull-ins. As you also might know from just checking the industry data, a lot of the capacity gets placed during the first half or first few months of the year, and our customers asked us to pull in.
As we now look into the calendar year, there are only a few months left, and the preparation has really now started for the ramps next year. This is why Doug mentioned that we expect Q3, our Q3 fiscal year, to be kind of the bottom, and then the business to pick back up again and accelerate going into Q4 and then into 2026.
Thank you very much. Just a quick follow-up. Could you also speak about the competitive dynamics for your SoC tester side, especially vis-à-vis your closest competitor in the U.S., and in the market such as GPU as well as the custom ASICs market?
Yeah, I think we are positioned very, very well right now. I would say that we should continue to hold our market position that we currently have. It's also very difficult for customers to switch test platforms in the middle of the ramps. For the foreseeable future, we believe we have a very strong market position and competitive position, both in the GPUs and custom ASICs.
Thank you very much. That's very clear.
Nakamura-san, thank you. Next, BofA Securities, Hirakawa-san, please go ahead.
You raised SoC test market for calendar 2025 from $4.5 billion at the midpoint to $6 billion, about 30%. Could you give us a sense of how you split this increase between the increase of quantity or number of chips and the increase of the testing time?
Thank you, Hirakawa-san. That's a good question, difficult question, I'd say. Before I talk about the test content or unit volume, I think I want to just go back and remind everyone that when we were putting back in January our forecast for TAM for 2025. This was right on the backdrop of the DeepSeek events, and there was a lot of uncertainty. When we reiterated that TAM in April, that was on the backdrop of the deliberation day, the tariff discussions that happened in April. Many companies at that point didn't offer any kind of market sizing. Many companies pulled their full-year guidance, and we didn't want to do that. We wanted to put our best estimate forward, even though there was a ton of uncertainty. Some of the 30% increase is a direct effect of us just being a little bit cautious given the uncertainties at the time.
That's probably at least half of the 30% was just the uncertainty factor. As far as test content or unit volume, I think it's probably, without trying to give a number, I would lean towards more of the unit volume basis. As there were adjustments in wafer starts or advanced packaging capacity that moved around, that would suddenly create some demand that was unforecasted, and we were able to respond to those demand fluctuations. Hopefully that helps you. Thanks.
Thank you very much.
Next, Jefferies Securities, Nakano myo-san, please go ahead.
This is Nakanomyo speaking from Jefferies. Can you hear me? Sorry, maybe my question is a bit technical. With regards to the upward revision of SoC TAM. SoC TAM is going to grow by like 40% at midpoint, maybe. Then your sales growth is only 20-some percent. What is the difference between TAM growth rate versus SoC revenue growth rate? What accounts for the discrepancy between sales upward revision versus TAM upward revision for SoC testers?
I think one thing that gets difficult is that our sales numbers we quote on a fiscal year basis and the TAM numbers we quote on a calendar basis. That often leads to some discrepancy. In our midterm plan, we have overall market share KPI of being greater than 58%. I can report that on SoC, we continue to be well above that metric. We're also well above that metric for the memory test. I apologize, but certainly the calendar versus the fiscal year creates some confusion. Yeah, sorry. Sanjeev also just mentioned exchange rate fluctuations because we show the TAM in dollars and our sales again. On a general level, our share in SoC has gone up. The memory tester, we remain relatively steady.
Next, Mizuho Securities, Yamamoto-san, please go ahead.
This is Yamamoto from Mizuho Securities. Can you hear me now? Apologies. I have a quick question. For FY 2026, can we expect an increase in sales next year? That is FY 2026.
Yeah, thank you, Yamamoto-san. I think I'll just repeat earlier without talking numbers, but we say we're very optimistic and we're seeing re-acceleration into 2026. We'll be able to talk more about that as we get further into the year and have a little bit more visibility. Right now, there's no reason to believe that 2026 is going to be a down year.
Just Q1, there is a very strong sell. Can we expect a similar big number in March 2027?
Yeah, I mean, the Q1 number, like I mentioned, was very special. Everything was aligning perfectly for us. Certain things that happened in this quarter are certainly things that we would hope as a company to be able to repeat long-term. It would be irresponsible for me to suggest that we can have a quarter this big in the future.
Yeah, of course. I know. Q3 would be bottom. From Q4, can we expect sequential growth toward the end of March 2027?
Yeah, I think it's fair to say that Q3 will kind of be the lowest part of the year and Q4 will come up. We would expect that trend to potentially continue into the first quarter of FY 2026. That's our expectation right now.
Thank you very much.
Yamamoto-san, thank you. Next, Citigroup Global Markets Japan, Shibano-san, please go ahead.
Hi, thanks so much for taking my question. My name is Masahiro Shibano from Citi. I have one quick question to you, Doug, about your 2027 outlook. I know it's really early to tell. However, you've showcased your very constructive outlook into CapEx in 2026. What do you make of the 2027 outlook based on what you are hearing in the street? Also, the question in the same context about the capacity, do you think the capacity you are adding right now is a permanent one, or is it going to be more like a temporary addition? Thank you.
Okay. Yeah, thank you, Shibano-san. As far as we look out beyond 2026 to 2027, the only thing that I would mention is that the areas that we look at as signals would be wafer starts and advanced packaging capacity that's being put in place by our customers. We also look at the end market, for example, the hyperscale capital budgets. In both those cases, I think there's been enough public discussion around continuation of the current trends. If those trends continue, then there's optimism that this growth cycle will continue beyond 2026 and into 2027. However, we know from our experience that there can be cycles that happen quickly. We do not assume that's going to be the case in how we run the company.
As to the production that we're putting in place, because we have a long-term belief in the industry, as I mentioned, getting to a trillion dollars in semiconductor revenue, we know that we're going to have to add this type of capacity in a permanent fashion. This would not be temporary capacity. In fact, even longer term, we probably will have to expand further. We have some of those plans that we have initiated already. Thank you.
Thanks so much.
[Foreign language]
[Foreign language]
Yeah, thank you, Shimamoto-san. Sorry. As far as why, the second question about why our sales deviated so much is mostly because of the pull-ins. If you were to spread some of the first quarter and even second quarter expectations across the year, it would look different. It is not that, as Sanjeev said, it is not that the second half is so bad. We just pulled from second half into the first half and even to the first quarter. As far as the other factors, as I mentioned earlier too, there were additional unit volumes that came into play that allowed our customers to ship more semiconductors that they did not even forecast. We were able to respond quickly to those scenarios and capture that business in the first quarter. Did I answer your question?
Thank you. The pull-in, the reason for the pull-in, did it have anything to do with the tariff impact or any political aspect to that pull-in? Is it just good demand and you being able to supply, meet that demand?
Yeah, I apologize for not answering that the first time. No, there was no tariff level. Factors really that presented themselves for this business. It was more just very large AI-related demand.
Understood. Thank you.
Shimamoto-san, thank you. Next, Tokai Tokyo Intelligence Lab, Kamisaki-san, please go ahead.
I'm Kamisaki from Tokai Tokyo Securities. My question is with regards to memory testers. Right now, I think HBM is driving your demand. Going forward, is it going to be big growth that will drive demand, or is it going to be HBM generation transition that will drive memory tester demand? Can you give color as to what will drive memory tester demand? Thank you.
Yeah, I think this is Sanjeev Mohan again. There are a lot of factors. I mean, it's a really good question. HBM is a growth driver for us, first of all. As the customers transition to HBM4 and HBM4E and higher stacks, that will drive more demand for us. Having said that, even though HBM is a growth driver, we have a very favorable product mix and dominance, particularly in the DRAM market. We are very optimistic about our memory business as well. We think the growth should continue. As the AI demand continues and the data center builds continue, that will drive more and more product volumes for the memory products as well. I think there should be a good growth opportunity for us.
Thank you. Understood. Thank you very much.
Kamisaki-san, thank you. We are almost at the end of the session. Maybe we'll just take one last question. Next, Aletheia Capital, Steven Liu-san, please go ahead.
I just have a question regarding your product mix in terms of the usage. We do know that before AI, wafer sorts and final tests were similar in terms of the ratio. Given the increasing usage of chiplets in chiplet architecture and more and more chiplets in the future, could you please share some color in regards to your ratio in terms of wafer sort demand as well as final test demand? As well, can you also share some color in regards to partial assembly handlers which you sell, as well as die-level tests and partial assembly tests? Could you just share some color regarding this? We do not need concrete numbers, but I think there should be a change in these sale trends. That is it.
Yes. I think you are hinting at it, and you're correct. I think the importance of wafer sort growth—these packages are very expensive. With the chiplet strategy that most companies are adopting now, the value of a known good die is increasing. We expect the overall business for the wafer sort will continue to grow, probably faster than what we would see at final test. As you also mentioned, the partial die or die-level prober, we also think that's a very exciting opportunity for us that we will be competing for.
Is that also a reason why you changed your reporting segments, moving the test handlers and the system-level tests into one big business? Moving on, is it okay to assume that for next year's growth, because we talked about sequential growth moving on from the fourth quarter, that's actually coming? That's actually also due to the increase of wafer sorts? Can we assume that?
Yeah. Hi, Steven. This is Doug. I think. Definitely there's going to be, as Sanjeev mentioned, there's going to be such an emphasis on known good die and testing at the silicon level. For sure, there's going to be new solutions that have to be developed. One of the key areas is to be able to do thermal control at that die or partial assembly level, because that's something that typically can't be controlled at a monolithic wafer level. That does represent additional upside. There is likely to be a change in the test flow at the wafer and die level, which could present some additional insertions. We see that as upside. That's not the reason that we changed the reporting segment. The reporting segment that we had before was a little bit antiquated with the mechatronics system.
We wanted to simplify it into kind of the capital equipment or the test system business, and then the service and others, more in line with what many other companies are doing in the industry. That's the reason for that. Very good questions. It doesn't mean that the final test and the first part of your question isn't going to go away. System-level test actually is an area where we see expansion as well. It's all additional content, additional insertions. One thing that we talk a lot about with our customers is this distribution of test and the ability to move test content around in the back end to allow the optimization of test content. We intend to play in nearly all of those insertions in order to service that distribution. Hope that helps. Thank you.
Yeah. Thank you very much. I'm done with my question. Thank you.
Thank you, Steven.
Steven, thank you for your question. I see other hands being raised, but we will conclude this session. Thank you for attending our first quarter FY 2025 briefing despite your busy schedule.