Asmpt Aktie

Good morning, ladies and gentlemen. I am Ben Poh, Head of Investor Relations and I will be moderating the call. On behalf of ASMPT Limited, welcome to our second quarter and first half of 2026 Investor Conference Call. Thank you all for your interest and continued support.

[Operator Instructions]. Before we start, let me go through our disclaimer. Please note that there may be forward-looking statements about the company business and financials during this call. Such forward-looking statements could involve known and unknown uncertainties and risks that could cause actual results, performance and events to differ materially from those expressed or implied during this conference call.

For your reference, the Investor Relations presentation on our recent results is available on our website. On today's call, we have the Group Chief Executive Officer, Mr. Robin Ng; and the Group Chief Financial Officer, Ms. Katie Xu. Robin will cover the group's key highlights for the second quarter and the first half of 2026 and provide outlook and guidance for the following quarter. while Katie will provide details on the financial performance.

Now I will hand the time over to our Group Chief Executive Officer, Robin. Robin, please?

Thank you, Ben. Good morning, everyone. Thank you for joining us today for the -- for our second quarter and the first half of 2026 earnings conference call.

Before we begin, I'm sure you're aware of the news that I will be stepping down as Group CEO on 11th August 2026. So this will be the last quarterly earnings call with all of you. I took over as Group CEO at the very height of the COVID pandemic in May 2020, helming the first investor conference call, which was Q2 2020, and it has been quite a journey. I am proud of what we have achieved as a business. I look forward to the company progressing on this transformation journey as it sharpens is focused on the backend packaging business. A business that has experienced rapid growth over the last few years. I'm grateful for your trust in me over the years as we clinical -- this journey together.

Looking to the future, I'm confident that the new CEO, Mr. Bassel Haddad will continue on this transformation journey and bring ASMPT to even greater heights. Given our technology leadership, strong foundations and great people. Thank you for your continued support.

Now, let me proceed with the earnings call.with some key business highlights for the first half of 2026. For the first half, I'm pleased to share that ASMPT achieved strong revenue and bookings driven by AI and the recovery in our traditional midstream business. As AI capabilities advance, semiconductor architecture requirements continue to expand beyond -- to continuous planning, would look orchestration, memory access and real-time data movement. These developments have placed increasing demands on semiconductor manufacturing to support both looks from data centers to each devices.

Next, SMT achieved record bookings, largely driven by strong demand from AI service. Data service continued to be a significant source of demand with accelerated adoption of SMT's high-flex I4 solutions, SMT bookings were also driven by demand for optical transceivers and China's EV segment. Semi AP bookings doubled year-on-year due to photonics and TCB.

This highlights ASMPT, multiple AP solutions, which are key enablers for present and future AI infrastructure. At the group level, our first half revenue increased 18.9% half of half and 42.5% year-on-year, mainly driven by mainstream and photonics. In Photonics, we saw customers ramping up production of high-speed optical transceivers for 800G and beyond. In mainstream, the group business experienced some recovery from traditional applications, such as consumer, industrial and automotive EVs.

Looking at our gross margin. This growth was driven mainly by SMT in Q2. And finally, we are pleased to report that our revenue and adjusted EPS bid consensus in Q2. Now let's look at the group business highlights. First, let's start with advanced packaging. This remains a strong growth engine for us in the first half with AP revenue growing 17% year-on-year and contributing 30% of group revenue. TCB, SMT high precision and photonic solutions were the largest contributors within the AP portfolio.

In TCB, we expect TAM to expand beyond USD 1.6 billion by 2028 due to growing AI investments and increasing packaging intensity driving greater demand for TCB solutions. In logic, other momentum for group's C2S solutions remains strong. This was supported by repeat orders for larger compound direct TCB tools from OSAT partners, one of the leading advanced logic customer.

In July 2026, the group received new back orders for more than 50 C2S TCB tools from OSAT customers. In CDW, we secured above order in Q2 2026 from a leading global IDM for advanced CPUs to support client computing and AI influencing. We also delivered ultra-fine TCB tools to the leading advanced logic customer.

In memory, the group continues to secure repeat orders from HMM manufacturers. However, the timing of some customers new to purchase decisions remains dependent on each mindful product rollout schedules. The group also entered into an exclusive joint evaluation program with a key memory player to establish its technology as a preferred production standard.

Next, panel-level packaging is emerging as a potential growth driver for ASMPT as the industry seeks to improve through book, scalability and cost efficiency. Beyond, we saw strong performance from our photonics solutions. Customers are ramping up production on high-speed optical transceivers for 800G and beyond as demand accelerates from bandwidth-intensive, low latency AI workloads. As a result, semi's trackable optical transceiver solutions revenue almost tripled year-on-year to approximately $75 million in first half 2026. The group is confident of further growth for the rest of 2026.

In co-optic package or CPO. The group continued to deepen its engagement with multiple leading global CPO players positioning well to gain market share as CPU adoption accelerates the group offers the most comprehensive range of CPO solutions, spanning ultra-high precision for card, TCB and hyper bonding. I will also elaborate more on this in the next slide.

Our semi mainstream business experienced significant growth supported by high utilization at leading IDMs due to improving demand for power management requirements for AIG centers and industrial applications. In China, revenue and bookings were particularly strong with wire bonding and dining to demand supported by ongoing AI infrastructure expansion and high wholesale utilization. Our SMT bookings hit record highs for Q2 and first half of 2026. AI service continued to be a significant source of demand with accelerated adoption of SMT high-flex high-force solution for large format bot is empty. And finally, beyond AI applications, I'm pleased to share that the group's mainstream business also experienced some recovery from traditional end applications, such as consumer devices, industrial systems and automotive is in China.

Let me now turn to Photonics, as shown on this slide. We have built a strong position across the entire photonics value chain, extending from pluggable optical transceivers to CPO. Our comprehensive portfolio spends SMT advanced packaging and ultra-high precision photonics assembly tools. In the current optical transceiver market, as shown on the left-hand side of the slide, as STD supports multiple critical assembly steps. Our SMT placement solutions are used for digital signal processor and passive components attached. Our high-end photonics and micro solutions enable precise placement of transmitters, receivers, chip on [ subno ] and optical accessories. Our solutions can deliver placement accuracy down to some micro level, a key technological differentiator.

Looking ahead, the industry is transitioning towards CPO solutions to address increasing bandwidth and power efficiency and latency requirements of AI infrastructure. SMT is well positioned in several key assembly steps. As you can see from the right-hand side of the slide, this includes EIC and PIC integration.through hybrid bonding and TCB solutions, optical assembly using flip-chip mass repo solutions and fiber attached Munich and micro lens assembly enabled by our MT class solutions.

We are excited about the potential in Photonics, which represents another important growth factor for ASMPT. With these highlights, let me now hand the time over to Katie, who will walk you through our group segment and financial performance.

Thank you, Robin. Good morning, and good evening, everyone. Let me take you through the group's financial performance. I would like to reiterate that Unless otherwise specified, the numbers I'll be referring to today are for the group's continuing operations only with adjustments made under non-HKFRS measures.

This slide covers our financial results for the first half of 2026. The group delivered revenue of USD 1.14 billion in the first half of 2026., representing an increase of 18.9% half-on-half and 42.5% year-on-year driven by semi and SMT. Group bookings reached USD 1.63 billion, representing growth of 68.1% half-on-half and 85.1% year-on-year. Both SMT and semi registered significantly higher bookings during the period. Book-to-bill was 1.43, the highest since the first half of 2021.

In the first half of 2026, group adjusted gross margin was 41.2%, this was 441 basis points higher half-on-half and 86 basis points higher year-on-year, driven by better gross margin from both semi and SMT Group adjusted operating expenditures was HKD 2.42 billion, up 2.4% half-on-half due to higher volume. It was also up of 15.4% year-on-year due to higher volume, unfavorable FX impact and strategic infrastructure R&D investments as we announced at the beginning of the year. Both adjusted operating profit and net profit improved half-on-half and year-on-year due to higher revenue and operating leverage.

In the second quarter, we delivered revenue of USD 630.0 million, which exceeded the upper end of our guidance. It grew by 24.4% Q-on-Q and a 52.1%, driven by both semi and SMT. Group Q2 bookings were USD 903.6 million, up 24.8% Q-on-Q and a 97.6% year-on-year. significantly better than anticipated for both semi and SMT, in particular, SMT's bookings came in much stronger despite a high base effect in Q1

Group Q2 adjusted gross margin was 42.5%, up 302 basis points Q-on-Q and 284 basis points year-on-year. Group Q2 adjusted operating profit was HKD 847.0 million, up 114.1% Q-on-Q and 268.8% year-on-year due to higher gross margin and operating leverage. Group's adjusted net profit was HKD 637.5 million, up 90.2% Q-on-Q and 253.9% year-on-year due to higher operating profit. Adjusted earnings per share was HKD 1.53.

Moving on to the Semiconductor Solutions segment for the second quarter. Semi delivered a revenue of USD 369.1 million, an increase of 34.9% Q-on-Q and 56.1% year-on-year. Q-on-Q and year-on-year growth was driven by photonics for AI-related applications and wide and die bonders for both AI and consumer-related applications. Semi Q2 bookings were USD 428.1 million, the highest since Q1 2022.

Bookings grew by 39.0% Q-on-Q due to wide and die bonders and photonics.and up 125.9% year-on-year due to wire and die bonders, photonics and a TCB. Semi's book-to-bill ratio in Q2 was 1.16 which has continued to expand for the past 4 quarters. Q2 adjusted margin for semi came in at 46.5%, up 10 basis points Q-on-Q and 150 basis points year-on-year. Q-on-Q increase was due to higher volume, offset by product mix, while year-on-year increase was largely due to higher volume. Adjusted segment profit was HKD 603.2 million in Q2, up 94.9% Q-on-Q and 170.1% year-on-year due to higher adjusted gross profit and operating leverage.

Next, let me move to the SMT Solutions segment performance for the second quarter of 2026. SMT delivered a strong Q2 revenue of USD 260.9 million, up 12.1% Q-on-Q and 46.9% year-on-year. It achieved record bookings of USD 475.4 million, up 14.3% Q-on-Q and 77.6% year-on-year. Both revenue and bookings growth were largely driven by strong demand from AI servers. Q2 SMT adjusted gross margin was 36.8%, up 551 basis points Q-on-Q and 429 basis points year-on-year, the highest since Q1 2024. Q-on-Q increase was due to favorable product mix and higher volume, while year-on-year increase was largely due to higher volume. Adjusted segment profit was HKD 284.8 million in Q2 2026, up 100.8% Q-on-Q and 386.1% year-on-year due to higher gross profit and operating leverage.

Now this slide highlights ASMPT's revenue breakdown by end markets. The computer end market was the largest contributor to group revenue at approximately 33% up significantly from around 10% in 2024, driven mainly by SMT Solutions, TCB and photonics applications. Please note that Photonics applications were reclassified from communications end market to the computer end market for both first half of 2026 and first half of 2025, reflecting the increase in alignment with AI-related applications. The consumer end market was the second largest contributor to group revenue at approximately 18%, mainly from group's traditional mainstream business. The automotive end market contributed almost 12% to group's revenue. Revenue growth was driven by EVs in China, while automotive for the rest of the world remained soft. The communications end market contributed around 10% of group revenue, mainly coming from high-end smartphone-related applications.

Lastly, the industrial end market revenue contribution was up marginally from 8% to 9% due to gradual recovery in industrial activity. Now as you can see from this slide, we are truly global business, partnering with customers across all major regions. China remained the largest market, contributing 42% of group revenue in first half 2026 driven by wire and die bonders. While share of revenue from Asia outside China declined from 37.9% to 36.2%.

Share of revenue from both Europe and Americas declined year-on-year from 23.7% in first half 2025 to 21.3% in first half 2026. The group continued to maintain a low customer concentration risk with the top 5 customers representing approximately 19% of total revenue in first half 2026. We have an existing dividend policy of distributing about 50% of the profit as dividend. For the first half of 2026 with adjusted EPS at HKD 1.94 for continuing and discontinued operations. The Board has recommended an interim dividend of $0.97 per share.

With that, let me now pass the time back to Robin for next quarter's revenue guidance.

Thank you, Katie. Let me now turn to our Q3 2026 revenue guidance. The group expects Q3 2026 revenue to be in the range of USD 630 million and USD 690 million. At midpoint, this represents an increase of 4.8% Q-on-Q and 46.3% year-on-year, exceeding current market consensus despite longer lead terms for certain materials.

Following a very strong Q2, the group still expects Q3 2026 bookings to grow by a high single-digit percentage sequentially. Mainly driven by TCB and photonics. The proliferation of AI will continue to drive structural demand for advanced AI applications and computing needs, benefiting the group's products. Our AP solutions are able to address complex technical requirements across the AI value chain, while our mainstream solutions can support the demands of extensive AI infrastructure bureau.

Beyond AI applications, we also see continued demand for some traditional mainstream applications. The group remains confident of revenue growth across both semi and SMT in 2026. Despite longer lead times, as mentioned above, and customers' dynamic AI product group schedules.

This concludes our second quarter and the first half of 2026 presentation. Thank you, and we're now ready for Q&A. Let me pass the time back to Ben to facilitate.

Thank you, Robin. [Operator Instructions]. May I request Gokul to unmute and raise your question.

Yes. Thanks, Ben. First of all, thanks, Robin, for your dealership and lack in your retirement. Maybe first question on TCB. It looks like memory TCB shipments are quite slow in first half of the year. Just looking at your Korea revenue momentum is down quite a bit year-on-year. So could you talk a little bit about when are we expecting the memory shipments to really start picking up meaningfully, especially given the order status seems to be still quite. So feels like it is very much a timing kind of gap? And is that the main driver for much stronger Q3 kind of momentum. And also on TCB, the 50 chip-to-substrate orders from OSATs that you highlighted. Could we also talk a little bit about the delivery schedule for the [ 50 tools ] how quickly do you expect to recognize that in revenue?

Okay. Thank you, Gokul, for your well wishes as well. Thank you very much. I think your first question is on memory, okay? And also a follow-up with what you expect TCB memory to pick up since the orders look muted, okay? Now let me answer your first question first. Now if you look at the industry as a whole, and according to what's out there in the market, the reports that are updated by the industry experts, it seems that HBM continues to be able to meet current requirements for GPU packaging. We also understand from this report that customers are also taking a little bit more time to make graded HBM technical specification. Now this obviously has implications on the product or status, which we also highlighted during our outlook as well. and definitely has an impact on customers purchase decisions for HBM TCB tools.

But having said that, from a home perspective, we continue to secure repeat orders from HBM manufacturers. Our stock to be one of these customers I'm also pleased to announce that we're also being deployed to [ HB4 ] high in high-volume manufacturing. At the same time, we're also having low volume manufacturing with multiple memory makers for His with our tools as well.

Last but not least, we also announced in the MD&A that we have entered into a JEP with a key memory player for advanced memory packaging particularly for HBM 5. So these are some of the development from a site that are very promising for HBM business. I think your second question is on TCB orders, right?

Yes. There are products that you said you are the TCB tools, any delivery...

Yes. Yes, Gokul, I think you understand that for TCB tools, typically the lead time a bit longer than the traditional tools, not counting the nonmain time for certain materials as well. So we believe some of these tools. Majority of these tools will be delivered probably in the first quarter of 2027. Some of these tools may be delivered in the second half of 2026. So we are trying -- I think the customers are trying to pull in and we are trying to meet the delivery commitment as much as possible. .

Got it. That's clear. Second question is on the photonics side, which is seeing very strong growth from pluggables. Can you talk a little bit about as we transition to CPO, how much value capture this ASMPT have, let's say, it's not one-on-one in terms of the comparison, but just to understand, like if we transition from, let's say, a pluggable base solutions to CPO-based solution how much of a value increase does SMT expect, given a lot of it is much more finer pitch. But at the same time, hybrid bonding is probably 1 area where you're a little bit kind of behind the market data. So I just wanted to understand how that venue capture expands as you go from pledgeable to CPO? And also, are you starting to see CPU-related orders also increase or most of the upside that we are seeing right now is paying rapidly from -- transceiver .

Okay, let me answer your -- probably the last bit. CPO is just more at this point in time. We believe from the packaging equipment standpoint, probably the reflection point will be somewhere between '27, '28, more like '28 for CPO. So most of the volume that we are experiencing right now to then what we call capable optical transceivers, which we also have a very strong position over there. Now in terms of CPO, we are confident, obviously, Micra is very well positioned for CPO market, but at the same time, even nice peak were Disney engaging very key customers in the CPO side of things. So when CPO picks up, as I said, probably '27, '28 in we're well positioned to capture the market going forward.

Now in terms of solutions, as we took some pain to show you what we are -- what we can anticipate on the CPO second, the number of applications, very key applications there, which our solutions are well fitted well suited for those very high precision placement kind of bonding requirements by our customers. So I think overall, I think we're competing of catching a good market share of the CPO market.

Okay. Thank you, Gokul, for your questions. And I'll move on to the next raised hand. I would like to request Sunny. Could you please unmute yourself and raise your question?

Yes, sure. Thank you and thank you, Robin, for all your leadership and contributions over the last few years. wishing all the best and exciting new chapter ahead. So my first question will be on the TCB orders as well. I want to double check on the current outlook for chip-on-wafer especially for the leading edge foundry side. And so how should we think about the visibility for the foundry client in terms of their migration to TCB for Q2 wafer from here? And also, how should we think about the competition for the opportunity? .

Thank you, Sunny. I think we said quite a few times that the way we view chip wafer in terms of opportunities. For the current year, we have been seeing that we demand would be still lower than cheaper substrate. But we have started to ship already chip-to-wafer solutions with AUR to the leading advanced or customer.

Now, the way we see it to wafer development is this. Now it all depends on the next generation of GPUs. We believe that the next-generation GPUs may go into chiplets. So that's where our TCB tools for tech wafer will be deployed. Because when it comes to chips integration, the die structure are very different from SOC. So for that reason, a chiplets kind of architecture require TCB for chip-to-wafer solutions. So we're looking forward for that to happen. And then when that happens, chip-to-wafer solutions or TCB solution will start to put up in a more meaningful manner.

Sorry. Let me quickly follow up, if that's okay. And so in terms of the potential opportunity from migrating to chap lot, do you think that a partum may still come through in 2027? Or do you think based on the current engagement with the client, that may come a bit later? .

They have been using our tools for testing, qualification and so forth. So we believe that it's a certain amount of urgency to start to use our T2 wafer tools. So we said before, I think 2027 could be the year whereby we see more demand for 2 wafer tools compared to .

Got it. And my second question will be on China. In China, obviously grew a lot in the last 12 months by about 50% year-over-year for first half of this year. So I wonder what's the key drivers for the strong growth? Is it still mostly coming from traditional or we start to see good progression on the advanced packaging side? And in the coming maybe 2 to 3 years, how should we think about the overall potential from China for advanced packaging given their very aggressive investment for leading-edge logic and memory. Do you think potentially China could be maybe larger than foundry, ID memory overall? .

Now let me talk about the mainstream, yes, we -- I think if you look at NDA, we have been assuming that the traditional mainstream business has -- we've seen recovery momentum. These are mainly coming from abandon the traditional viable and Timon coming from China side. We also -- on the end application side for these tools, they are typically consumer-driven EVs and so forth, right? So the China for sure, the China ministry business is on the recovery path. Now in terms of advanced packaging, they are still lower than the rest of the world, but we believe that this momentum in China will continue to increase over the yes.

Thank you, Sunny. So next, I think Daisy's hand is raised. So I would like to request Daisy to unmute and raise your questions.

And congrats on the strong results in Robin, we show all the best for your retirement. My first question is regarding hyper bonding. And you mentioned that hyper bounding has moved to their sampling stage. Could you share us with possible timing that when you expect that it could be qualified for the memory and the logic customers, respectively. .

Yes. Thanks, Daisy. Now hybrid bonding, yes, we are confident of our hybrid bonding solution. As we have highlighted before, we believe our second-gen generation HP solutions are very competitive, I'll say, in terms of certain key metrics like alignment, precision, bonding accuracy, footprint efficiency and also UPH. We still of the view that hybrid bonding inflection point could be still -- from our perspective, of course, a couple of years down to growth may be '29, '30 kind of time frame or even '31. So it's very dynamic. So we can't really put a definitive time frame on hybrid bonding yet.

But what we do know on the flip side for TCB is that is possibility that the memory chip hire, the step hype for HBM, there's a possibility that it could increase from the current S5 micron to beyond also. So if that happens, then TCB could be -- TCB could be extended for packaging beyond he's been for into [ 5 ], right? So that's sort of one factor to take into account the deployment of hybrid bonding solution.

Now as far as hybrid bonded solution for ourself. This year, you were definitely trying to break into the top hybrid bonding customers. if we can break into that space, I think call position us well hybrid solution for the times. Yes, indeed, we are doing same thing for both logic customers as well as Asian customer for a funding solution, yes.

And my second question is for Katie. So this quarter, we see that both S&P and semi Solutions, the margin in clock. And we all know that this year, both the OSAT and the TCB companies CapEx is very strong. So again, how should we think about the normalized margin? And also, how do you see these CapEx spending momentum? Will it continue into 2027?

Then maybe you'll break your question in 2 pieces. One is on gross margin, the other 1 is more so I think on the top line side due to a -- so let me answer the first question. In terms of gross margin, we announced in Q2 already. You, guys, can see the numbers going kind of looking ahead, right, looking to Q3, we actually -- we think the -- we believe the gross margin will actually remain quite stable for both businesses. the semi side, it will continue to benefit from the favorable product mix of TCBs for China, high NDAs, et cetera. while it will be moderated by win mix, right? So that's semi side.

On the SMT side, it will continue to be in mid-30s in terms of gross margin at this volume level, right? So I think to put these 2 together, as we look out the Q3 and maybe the quarters out that the overall, the group margin could stay above 40%.

The next question, I think that you also asked about the CapEx spend to the top line impact rate. I'll start and if Robin have anything to add. I think overall, I think the -- yes, the CapEx has been spending strong and we definitely have experience that in our actual results. And I think going forward, this momentum will continue to the line of sight that we have for the remaining of the year. And we -- I think we touched on that in our sort of the outlook for the MD&A, yes.

Okay. Thank you, Daisy. Next, I would like Leping to unmute to raise your question.

So my first question is about the logging optical transceivers. You mentioned that your revenue tripled in the first half of this year. So how we should look the sustainability or the potential of this business. Will you become even bigger business versus your other, like the TCB or other high potential business.

So yes, I think Photonics is really interesting. So if you look -- if we look at the Q3 optic, the relay out that it's going to be increased by high single digit Q-on-Q, part of the increase, I would say, is also due to photonics. So we believe our photonics okay in Q3 will continue to increase sequentially as well. right. So I think it's very sustainable. So I think we probably give you an idea that the photonic, we believe the photonics business in the second half or the demand in the second half will be growing as well relative to the first half.

Now optical transceiver definitely is part -- is a bigger portion right now for photonics. As I mentioned earlier, CPO is a smaller portion. In terms of TAM, I think you asked something about TAM, right? So let me give you some color on the TAM. I think for this call, we'd rather not focus on a specific 10 number to date. But as I said, I want to emphasize that we will continue to see a meaningful growth runway in both trackable and see CPO photonics in years to come. I think more importantly, I think the opportunity for us is really expanding as customers move to a more advanced optical architecture and we strongly believe that we have a pretty early to participate across several critical process steps in that particular value chain.

Now I also want to be mindful and I also want to be careful about drawing direct comparison with TCB at this stage for Photonics. As you're aware, TCB is a much more mature market. with much more clearer visibility in terms of adoption and market sizing. Photonics, particularly for CPO is still in the earlier stage of development. So I think what's important for us today is that we continue to see and see strong customer engagement on our part. We see increasing adoption activities, our tools. And we see a lot of opportunity across multiple session participate both in the CPO -- more so it's in the optical conceivable booking.

And also, the second question is about your third quarter guidance. You mentioned that the -- you see strong growth -- sustained growth despite a long lead time of certain materials. So number one is, how -- what are the longtime supply chain issue you are facing? And it seems to be the third quarter the revenue, the guidance is slightly slower than the booking is slightly slower than the second quarter booking growth. What was the reason? Is it because you are facing some supply changes or you are still seeing -- and what's the growth coming from? It's coming from mainly the SMT or coming from the SMT .

Okay. Good. The rapid -- let me answer the only time question. On the rapid investing growth has certainly led to supply chain tightness, resulting in longer times for certain materials, which affected both the SMT as well as semi operation. I think you alluded to also the probably the longer conversion time from bookings to billing. So definitely a long noted time is one factor.

Now the other factor that you might have to take into consideration is that the increase in our semi product mix towards advanced packaging also contributed to the elevated auto to building conversion time because AP has relatively longer manufacturing lead time compared to our traditional business. So for this couple of reasons. The group's booking conversion time is estimated to be a little bit more extended before to be around 6 to 9 months, roughly, depending on the products itself.

Thank you, Leping, for your question. And next, I would like Kevin .Kevin, could you any and raise your question?

And thank you, Robin, especially for your past leaderships, wish you all the best in your upcoming retirement. Now my first question will be on the TCB TAM. I think in the announcement, we mentioned that our TAM right now, we're expecting that could go beyond USD 1.6 billion by 2028. I was wondering what do the management see this upside coming from? Would it be more coming from memory or logic especially given right now there are some growing market concern over the memory outlook. What do we see any changes for memory sector driving the TCB outlook. Would it be any additional customer coming in?

Thanks, Kevin, for your well wishes as well. No. Yes. I will not give you any specific because this is a half year. We aim to provide an official update at this year's full year earnings call on the TCB TAM. However, having said that net may try to provide some color on the TCB TAM situation. Yes, indeed, we expect our TCB terms to expand beyond USD 1.6 billion by 2028. That was the last time we teach-out this number during our Q4 earnings call.

Now this -- we believe this expansion of TCB TAM by 2028 will come from several areas. First, there will be increased their and CPU demands to support the growth in AI influencing needs in better maybe agency down the road. The emergence of this influencing or agent AI is driving a shift from GPU-centric design. There was a more balanced CPU to GPU configurations, basically increasing CPU or increasing CPU content for AI growth. So as agentic AI requires more autonomous planning, continuous background processing and so forth and execution of complex local motors. We believe the demand for high-performance CPUs and greater memory bandwidth and capacity is accelerating, right? So this will contribute to more TCB demand down the road basically with the increase in the number of CPU that need to be packaged.

Now I think alongside this, you can imagine, we believe the HBM volume also increased alongside this kind of development.

Now the second reason we can gain on is we see increased log packaging requirements for cohorts as well in the years to come. And last but not least, panel packaging, including embedded and even a little bit of bridge tie kind of solutions where we are participating will also contribute to increase in TAM for TCB by 2028. Now we've also, last but not least, we also believe this expansion is expected in spite of the industry-wide phenomenon from continuous improvement in HBM TCB-UPH. So I think this sort of gives you a color driven of why we're confident that the TCB TAM will expand beyond USD 1.6 billion by the time we're in 2028.

Right. My second question actually I want to go back to hybrid funding. Just wondering for our Gen 2 tool, is that mainly for wafer-to-wafer or tied to wafer capable. And going forward, I know that we -- I think we are expecting inflection point probably sometime 2029 or 2030. What do we see could drive the demand either if we're pulling the demand a little bit earlier than the time we expected.

Now is that wafer -- if you're talking about wafer-to-wafer bonding, we do not have the triple -- solution in-house. Now in terms of inflection point, it all depends because it's -- the other side of the coin is really TCB, right? So you know that the TCB in terms of technology, we have been advancing TCB technology together with our customers as well. So it depends how TCB advance and also the market development in terms of is in particular.

Now as I mentioned earlier, for his if there is a relaxation of the stack high then there's a possibility that we could continue to extend the use of TCB beyond HBM 4 into HBM 5, and that potentially can push back the adoption of [ hundred volume ] for HBM stacking. This is suppose our view -- so I hope I sort of give you an indication why we continue to say that the HP, we believe the HP adoption will probably come in the kind of time frame.

Thank you, Kevin. Thank you for your questions. And next, I would like to invite Tracey to unmute to raise your question. .

Thanks for the opportunity to raise questions, and thanks, Robin, for your leadership and wish you all the best ahead. My first question is about -- so in China market, we see both advanced logic and the memory capacities are having fast expansion. So just wondering what is the revenue order scale of your TCB business in China currently? And how you expect its growth? .

Yes, Tracy, I think you mentioned, but it's probably worth repeating as well. Yes, I think we -- China in terms of advanced packaging, China is relatively small than rest of the world at this point in time, but we believe with all this advancement in terms of advanced packaging in China, we believe that the trajectory for AP in China looks interesting in the years to come.

Okay. I got it. So maybe my second question is about your bookings guidance for the next quarters. Are there any bleed between semi business and SMT business for your bookings outlook. Let me give you -- for Q3. I hope you -- right? So bookings for Q3 this year is expected to grow as highlighted by high-single-digit. Sequentially, mainly driven by semi products, TCB and photonics in particular. Of course, we just announced this morning that we won more than of more than 50 TCB, C2S tools from OSAT customers that would certainly contribute to the increase in terms of future bookings Q-on-Q for semi.

Now SMT bookings on the underhand likely to moderate or decrease due to a very high base already effect in Q2. But certainly, we still believe it's an elevated level. So on a year-on-year basis, SMT booking would still be higher, driven by strong demand from service.

Okay. Thank you, Tracy. I think we have time for just 1 final question. Oh, no, sorry. There's no more raised hand. So I think that will be it for our Q&A session.and I will pass the time back to Robin for his closing remarks.

So thank you all for all your well wishes before we end, let me capture some key takeaway from today's discussion. First, ASMPT continued to experience strong revenue and bookings driven by AI and traditional ministry application, following a very strong Q3 -- Q2 financial performance revenue guidance is above market consensus. Second, advanced packaging remains strong growth in [ NGS ] with TCB, SMT high precision, and photonic solutions, the largest contributor within the CP portfolio. In particular, Photonics represents a compelling growth opportunity for us. Overall, we are well positioned as we enter second half of 2026.

So ladies and gentlemen, with that, this concludes my last earnings call at ASMPT. This is my 25th such call, and it is like a nice number to sign off. As I enter the next phase of my customer journey in a couple of weeks, I hope that you will continue to support ASMPT. I look forward to catching of with you some of you in the smaller group meetings on this trip. So thank you once again for joining us today. Goodbye.

Good morning, ladies and gentlemen. I'm Ben Poh, Head of Investor Relations. And today, I will be moderating the call. On behalf of ASMPT Limited, welcome to our First Quarter 2026 Investor Conference Call. Thank you all for your interest and continued support.

[Operator Instructions] Before we start, let me go through our disclaimers. Please note that there may be forward-looking statements about the company's business and finances during this call. Such forward-looking statements could involve known and unknown uncertainties and risks that could cause actual results, performance and events to differ materially from those expressed or implied during this conference call. For your reference, the Investor Relations presentation on our recent results is available on our website.

On today's call, we have the Group Chief Executive Officer, Mr. Robin Ng; and the Group Chief Financial Officer, Ms. Katie Xu. Robin will cover the group's key highlights for the first quarter 2026 and provide outlook and guidance for the following quarter. Katie will provide details on the financial performance for the quarter.

Now I will hand the time over to our Group Chief Executive Officer, Robin.

Thank you, Ben. Good morning and good afternoon and good evening to all. Thank you for joining us today for our first quarter 2026 earnings conference call.

Now let me start with the key business highlights for Q1. This quarter, I'm pleased to share that ASMPT achieved the highest quarterly bookings and billings in the last few years. We continue to see AI drive demand across multiple products as the rapid evolution of AI increases the value and complexity of back-end semiconductor manufacturing. New AI architectures demand heterogeneous integration, tighter interconnect pitch, higher bandwidth and power efficiency. And these requirements are driving higher precision, alignment and process control needs across packaging flows, benefiting a wide range of the group's product from TCB photonics, CPO, flip chip and mainstream volume and die bonding and pick-and-place solutions.

Let me provide some color on these specific product areas. First, let's look at TCB. In Logic, we delivered sizable shipments for chip-to-substrate applications, reinforcing our leadership in chip-to-substrate TCB. We received bookings for 4 ultra-fine-pitch chip-to-wafer TCB tools, featuring our fluxless plasma-based AOR technology from a leading advanced logic customer. We are also actively engaging key logic players across multiple programs, and we are well positioned for more opportunities as the industry advances towards more complex logic chip architectures.

In memory, our TCB tools remain at the forefront of technology development. A key memory player is using a flux-based TCB tool for assembly and this customer is also qualifying a Fluxless AOR solution for HBM4 16-high.

Next, we'd like to share an update on Photonics. I'm pleased to report that our Photonics revenue grew nearly fivefold year-on-year, benefiting from strong demand for high-speed optical transceivers of 800G and above. In addition, our 1.6T transceiver solution received bulk orders from leading optics suppliers in the data center networking supply chain. This demonstrates strong traction for our optical transceiver solution as the market leader.

I would like now to touch a bit on co-packaged optics or CPO before we move on to the next item. CPO represents a paradigm shift in AI system design, bringing optical engines closer to compute silicon to reduce electrical losses, lower power consumption and improved system efficiency. Our CPO solutions enable high precision bonding to integrate diverse components, including fiber array unit, microlens, electronic IC and photonic IC into a single high-performance optical engine.

We have deepened our engagement with multiple leading global players and this positions the group well to gain market share as CPO adoption accelerates. Looking now at our flip-chip solutions. I'm pleased to update that they registered strong bookings growth, both Q-on-Q and year-on-year.

This momentum is coming from two areas. First, there is an accelerated adoption of 2.5D packaging for larger AI package sizes that is driving a steady pipeline of opportunities for embedded bridge die-bonding solutions for both chip-on-wafer and chip-on-panel solutions. Second, we also gained traction in panel level fan-out for radio frequency and power devices. Both these areas are well solved by our flip chip solutions, which combine cost efficiency, scalability, high placement accuracy and strong throughput.

And finally, in our mainstream business, we registered strong bookings for both SEMI and SMT. SEMI's mainstream business benefited from sustained utilization at leading global IDMs and OSATs alongside rising demand for AI data center power management solutions. In China, there was increased demand for wire and die bonding applications. For SMT, we achieved record booking gains in driven by strong customer demand across AI servers, optical transceivers and China EVs. In particular, SMT's high-flex high-force solutions for large format boards are a leading choice for AI server assembly.

As we broaden our AI customer base, we are fully committed to delivering the highest quality of solutions and services. ASMPT was recently recognized with a prestigious Intel Epic Supplier Award for 2026, the highest supplier recognition award for excellence in business collaboration. This is a reflection of our strong technical capability and deep engagement with our customers.

With these highlights, now let me hand over the time to Katie, who will walk you through our group and segment financial performance.

Thank you, Robin. Good morning, and good evening, everyone. Before I start, I would like to say that unless otherwise specified, the numbers I'll be referring to today are for the group's continuing operations only, with adjustments made on the non-HKFRS measures.

This slide covers our group financial results for Q1 2026. In Q1, the group delivered revenue of USD 507.9 million flat Q-on-Q, but up 32.0% year-on-year, driven by SMT and SEMI. Group revenue came in above market consensus and was the highest in the last 3 years. Group quarterly bookings exceeded expectations with SMT bookings at a record level. Group's booking reached USD 727.0 million, up 46.0% Q-on-Q and 71.6% year-on-year, the highest in the last 4 years. This strong growth came from multiple products, notably SMT products, wire bonders and die bonders and photonics.

Group adjusted gross margin was 39.5% Q-on-Q, up 357 basis points due to higher gross margin and revenue contribution from SEMI. The year-on-year decline of 151 basis points was due to a higher revenue contribution from SMT.

Group's adjusted OpEx declined 4.6% Q-on-Q but increased 12.4% year-on-year, largely due to unfavorable FX impact and from strategic infrastructure and R&D investments as we have guided for 2026 during our last earnings call. Both adjusted operating profit and adjusted net profit improved Q-on-Q and year-on-year due to higher revenue and operating leverage. Adjusted EPS was at HKD 0.81, up 118.9% Q-on-Q and 189.3% year-on-year, which was above market consensus.

Moving on to the Semiconductor Solutions segment. In Q1, SEMI revenue delivered USD 274.5 million, up 12.2% Q-on-Q and 14.6% year-on-year. Q-on-Q, growth was driven by high-end die bonders and TCB, while year-on-year growth came in from multiple products, largely driven by AI-related applications.

SEMI bookings were USD 309.6 million, up 22.6% Q-on-Q and 43.2% year-on-year due to higher demand for wire bonders and die bonders driven by China OSATs, high-end smartphone-related applications AI-related power management applications and optical transceivers. SEMI achieved a book-to-bill ratio of 1.13, which marks 3 consecutive quarters of improvement.

SEMI adjusted gross margin reached 46.4%, achieving the guidance we set last quarter. Adjusted gross margin improved by 594 basis points Q-on-Q but declined slightly by 37 basis points year-on-year. The significant Q-on-Q improvement was mainly driven by high volume and favorable product mix. SEMI adjusted segment profit was HKD 309.4 million, up 165.9% Q-on-Q and 16.8% year-on-year. The strong Q-on-Q improvement was mainly driven by higher gross margins and operating leverage.

Let me move to SMT. SMT Q1 revenue was USD 233.5 million, down 11.0% Q-on-Q but up 60.7% year-on-year. Q-on-Q decline was due to seasonality, while year-on-year increase was due to strong demand from AI servers and China EVs.

As mentioned earlier, SMT achieved a record bookings of USD 417.4 million, up 70.0% Q-on-Q and 101.1% year-on-year. This was primarily driven by strong demand from AI servers, optical transceivers and China EVs together with robust China demand arising from global data center expansion. SMT adjusted segment profit was HKD 141.8 million, down 28.3% Q-on-Q due to lower volume, but it improved year-on-year.

Now let me hand the time back to Robin for the outlook and the revenue guidance.

Let me present our Q2 2026 revenue guidance. The group expects Q2 2026 revenue to be in the range of USD 540 million and USD 600 million. At the midpoint of USD 570 million, this represents an increase of 12.2% Q-on-Q and 37.0% year-on-year. Notably, our midpoint revenue guidance exceeds current market consensus, and it will be mainly driven by SEMI.

Group bookings in Q2 2026 are expected to remain elevated for both segments, though SMT will be down Q-on-Q due to the high base effect from Q1. The continued proliferation of AI expected to drive structural demand growth for both SEMI and SMT in 2026 across multiple products. This includes our flagship TCB and HP solutions, photonics and CPO to mainstream wire and die bonding and pick-and-place solutions that enable AI infrastructure deployment. Looking ahead, structural AI-driven demand is expected to support revenue growth across both SEMI and SMT in 2026.

This concludes our first quarter 2026 presentation. Thank you, and we are now ready for Q&A. Let me pass the time back to Ben to facilitate.

[Operator Instructions] And I see a raised hand from Gokul of JPM.

Great results, and also thanks, Robin, for your amazing leadership over the years. So first question is on memory TCB. What are we seeing in terms of bookings and potential for bulk orders for memory TCB given that we haven't really seen any big bulk orders in the last maybe couple of quarters now. And at the same time, the R&D progress seems to be quite good on both flux-based and fluxless. And in addition to that, could you also talk a little bit about your engagement with the bigger memory vendor that the market is talking about, which has largely been using internal TCB tools? Do you see that there is an opportunity opening up with this customer, which will definitely expand your addressable market?

Thank you, Gokul. I think there are a number of questions within your questions. Let me address I think the first one first. You're asking about memory TCB bookings and potential for bulk orders for TCB and in the last few quarters. The last bulk order that we received was in Q4 2025. We're still confident that we are well placed to -- well positioned to receive orders from memory makers as far as they are ready to dish out orders for equipment supply ourselves.

Now I think your second question is on fluxless, R&D fluxless, how is it going. I think we are making good progress, especially on the logic side, as we have mentioned, we have won 4 tools in Q1 for CoW fluxless applications. We believe this is a start of this particularly exciting program. I think as the industry continue to migrate to more complicated GPU or even ASIC architecture, I think at some point, they may have to switch to a chiplet kind of configuration rather than using SoC, and that's where the opportunity to use our TCB fluxless tool for chip-on-wafer application will be there.

So since now that we are already in that supply chain, I think, again, I think we are really well positioned to capture more opportunities for CoW fluxless application going forward.

I believe your third question relates about the biggest memory player who is used to using internal TCB tools. Yes, I think we are unable to really name or confirm any specific collaboration with any customer, I hope you understand. What I can say that in the process of finalizing an evaluation program with a key memory player. We definitely see this as a positive step, possibly in the future, enabling our group's technology from memory as a process standard in the future. So we are excited about this potential collaboration going forward.

Gokul, very quick. Just going back to the question on the memory TCB bulk order. I just want to -- probably to say that we want to reiterate, right? Our TAM forecast is actually of the $1.6 billion is intact. And you're probably reference into some of the recent adjustments for the next generation of GPU HBM4 road maps. They are leading to some -- maybe quarter-to-quarter, there might be some variability in the times of a customer's decision. But our activity level remains very healthy, but it could be uneven from quarter-to-quarter.

Got it. That's very clear. Second question I have is on your photonics, obviously, very strong growth, 5x kind of growth. Could you help us size the like photonics business, I think you've said it long back that it was probably under $100 million annual run rate ballpark? Or is this still -- is now much bigger than that?

Yes. I think, Gokul, you...

When we transitioned?

Yes. Maybe answer the photonics first, Gokul.

Well, go ahead, Robin. I'll follow up later.

Yes. There's a little bit of feedback on this.

Yes. Your line is breaking up, actually, Gokul.

Okay. Anyway, I'll answer the Photonics question first Gokul. Yes, Indeed, when we look back in a deep dive into the photonics and the CPO market recently, actually, the TAM looks even more promising than before. right? So -- but we are not ready to disclose the TAM at this point in time. But I can tell you the TAM looks bigger than before. So when we combine the optical transceiver TAM and the CPO TAM, I must say that the TAM looks interesting. We are definitely paying a lot of attention in this area.

And fortunately, I think for photonics, we are already a very strong player in the optical transceiver market. And I think for the CPO, you probably will follow up with your question on CPO as well. I think we're well positioned with some key players already. Our solutions, I would say, have been designed in with a number of key CPO players. So when the adoption takes place, I think we're in a good position to capitalize this opportunity for CPO. Maybe back to you for second question, please.

Yes. So I think just to follow up on that CPO comment, Robin, the market understanding, obviously, is that hybrid bonding plays an important role in CPO for the EIC, PIC attach. And obviously, your hybrid bonder is still in qualification, especially the second generation. So could you talk a little bit about the progress there? And maybe also help clarify. I think there are multiple die attach steps, not just hybrid bonding, I think beyond that as well. So I just wanted to understand like what is the span of like SMT's involvement when it comes to CPO, just beyond the EIC, PIC attach?

That's right. I think we also said in the MD&A, we are participating in several key high-precision bonding areas for CPO, one of which is FAU attach on PIC. The other one is what you mentioned, stacking EIC on PIC. The third application we can think of is micro lens attached on PIC and also finally, the whole optical engine on the substrate. So we have solutions actually for all these key bonding solutions.

That's why we feel particularly quite excited about the CPO market. And as I said earlier, I think the TAM looks interesting, maybe not in the initial years. But I think the CPO will accelerate probably from '28 onwards, and the TAM in fact, looks very interesting for CPO. So these are the areas we are participating.

Now back to your question on EIC on PIC, hybrid bonding is one solution. I think CPO players are also exploring whether they can use TCB for the application as well. So if they use PCB, that will be also very an interesting market segment for us.

I'm seeing a next raised from Daisy. Daisy, could you please unmute yourself and raise question?

Yes. And my first question is regarding the OSAT CapEx. So we saw that the OSAT CapEx is getting higher and higher for this year. So which area do you expect to record the highest growth for the year based on the current order visibility? And I mean the regions. And also, China is the largest revenue contributor last year around 41% of your total revenue. Within China, do you see that still advanced packaging, I mean your TCB and other hybrid bonding tools growth, outgrow the mainstream or mainstream for this year is also very strong. That's my first question.

Thank you, Daisy. In terms of -- I think your first question first in terms of OSAT CapEx yes, in fact, we are experiencing strong demand, I would say, on the OSAT front, mainly coming from wire bond die bond because of the AI demand for infrastructure. I think we have been talking about this for a few quarters already, but in Q1, the demand is particularly very strong. What is driving this is really power applications that go into data center. So this require new power devices and because of that new capacity is required.

So that's driving a lot of our wire bond and die bond and also not forgetting SMT as well. We mentioned that SMT had a very fantastic booking in Q1 highest so far in history, largely also driven by AI server boards. And in there, there are a lot of power packages using tools from SiP, tools from SMT as well. So all this infrastructure deployment and spending are driving a lot of our mainstream tools, both in SEMI as well as in SMT. Now I think your second question is about China, whether China region, whether AP grow, is it faster than mainstream or mainstream is still very strong.

I think typically, in China, say especially on the SEMI side, the mainstream side are definitely stronger than the AP side. However, on the SMT side, in the rest of the region, still stronger than the China side. So there's a mix in terms of China demand coming from SEMI and SMT segment.

And my second question is also regarding the optics, photonics solution. So you mentioned that the revenue delivered fivefold increase year-on-year this quarter. And may I ask why we suddenly saw a very strong pickup in this segment. And I believe you mentioned that regarding the booking, the photonic solutions is both in your SEMI solutions segment and SMT segment. May I ask what tools for the SMT and what tools within the SEMI solution?

Yes. I think it's really all AI-driven, data center driven, as you can imagine, as the industry continued to increase the silicon compute, the transmission side has to match that capability as well. So that drives a lot of growth presently in terms of optical transceivers and the industry is moving from 400G to 800G to 1.6T, and we have a very, very good solution for optical transceivers.

This segment has been seeing steady growth for a few quarters already. We have been reporting there. So it's nothing new to us. We have been saying that optical transceiver is a good market for us. We have been quite dominant in that space. We've been winning market share as well. So that's something that we are experiencing from many quarters already in terms of photonics.

Now your second question is for -- both segments indeed are participating in this area. Now for SMT, mentioned in the optical transceiver there are many, many components, some require higher precision than others. So for those components that require higher precision bonding. They use our SEMI tools for that purpose. For those that do not require a lot of precision they use our SMT pick and place tool to bond those components. So both segment SMT and SEMI are actually benefiting from this surge in demand for optical transceivers.

Yes. Next, I will request Kevin from Citi to unmute.

So I have two questions. Number one is that I would like to get some more detail on the booking guidance outlook. As you see right now, our booking is back to -- especially like SMT back to a record level. So can we get for the coming quarters, do we see -- have a rough sense of breakdown for booking into the, say, SEMI and SMT and specifically, which region are we seeing the most growth from. And also, last time we mentioned we're seeing some improving visibility. Is this still the case so that right now, approximately how many months of visibility do we have right now? .

Thanks, Kevin, for the question. Now I think you're referring to Q2 bookings. Now we -- as you know, we don't really guide but we can definitely give you some color we see bookings in Q2 this quarter to remain elevated. Booking may, however, moderate Q-on-Q but still expected to grow strongly on a year-on-year basis. We believe in large part, we continue to benefit from the secular demand for AI-related applications. And in the infrastructure spending, plus, of course, overall improving market condition as a whole. Giving a little bit of color as to the or the segment booking.

Now for SEMI, Q2 SEMI bookings are expected to increase Q-on-Q and more significantly higher on a year-on-year basis. However, for SMT bookings are likely to decrease Q-on-Q due to high base effect as we have reported, Q1 booking for SMT was at a record high. So we don't expect the current level to continue on a Q-on-Q basis. However, having said that, SMT bookings are also expected to be higher on a year-on-year basis. Now you asked about visibility.

Looking ahead, while we are definitely more optimistic about business compared to some quarters back, there is less visibility for the second half of 2026 for the whole group. I think this is pretty normal for a business that we can't really look too far away. So I hope I answered your question, Kevin.

Yes. Maybe real quick, I think Kevin was asking about the booking by region. And Kevin, we actually sort of answered it already when we were talking about -- when Robin was addressing Daisy's question. Since overall, the regional mix will stay relatively stable. But like what Robin was mentioning about the strength of China OSATs. So there will be a little bit more booking from China. But overall, it's quite steady.

Great. My second question is on the EMIB outlook. I think recently, there has been some demand pickup on this technology. I'm just wondering what type of or tools are addressing this kind of demand? And also our position to the share allocation in this type of technology. Do we need a special type of TCB and that require customization as well?

Yes. I think, Kevin, I can't hear you properly. I think you mentioned EMIB-T right? .

Yes.

Okay. So A couple of layers we have to understand on the EMIB-T program. If you are talking about embedded die bonding, we believe this is on a large substrate, probably 510 x 500. Unfortunately, TCB, we are not ready for that yet in terms of that kind of panel size because it will take some time for us to deliver a tool of that size for TCB. But however, if this EMIB program takes off and we believe it will, we are already -- this particular customer is really using our tool for CoW application.

So do you have to -- I mean, if this program proliferates, right? So they will probably need also more tools to place a lot more components on the EMIB-T substrate. So I think I think we will benefit from that particular area that means on the CoW tools, which we're already in. But if you're talking about embedded die bonding for the EMIB-T, we are not there yet.

Next, I will move to Sunny. Sunny, could you please unmute and raise your question?

Hello. Could you hear me okay?

Yes. Very well.

Congrats on the very good results and thank you, Robin, for all your leadership over the years. And so my first question is on your opportunity on the logic, especially on chip-on-wafer. And so Robin, earlier, you mentioned the chip-on-wafer migration for TCB could be somewhat related to chiplet. That was a bit surprising to me because I always think that the chip-on-wafer migration for TCB should be related to larger package. And so how should we think about from here? You just secured 4 tools from a leading-edge foundry customer. How aggressive are they in migrating to TCB for chip-on-wafer from here? How should we think about when you may get another bulk orders? Would that be in second half? Or will you need to wait until maybe 2027?

Okay. Thanks, Sunny, for your question. Now when we -- when I mentioned about chiplet, between chiplet and SoC, when the GPU architecture is using SoC, there is less need to use TCB to place the SOC onto the interposer because you don't need that kind of precision. But when you try to -- when you go into SOIC kind of packaging, you need more precision to put those chips onto an interposer. So that's why TCB will be needed going forward. So I think as the industry migrates from an SoC structure to an SOIC structure, we see increasing use of TCB for that application.

So however, having said that, we have been telling you guys that for 2026, the number of tools for CoW will not be significant because it all depends on the migration to the next chip architecture. So we believe '26 will not be high for CoW. But going forward in the years to come, I think there's a meaningful TAM over there for CoW application for logic. Now how aggressive is this migrating to CoW? Yes, I think I already answered your second question as well. So '26 will not be high, but '27 would be meaningful.

Also, if I may follow up on my first question. Also, these leading-edge logic customer, they are already working on the follow-up solution beyond CoWoS, meaning CoPoS. So yes, so from your perspective, for their CoPoS, they will start from smaller form factor, 310 x 310. And so from your perspective, are you seeing any signs of clients trying to pull forward the technology development. And for CoPoS, how should we think about your overall opportunity, especially around chip-on-panel.

Yes. I think we are -- as we speak, we are developing tools for CoP. So we deal to deliver demonstration tools sometime this year. So that part, we are already engaged with the key advanced logic customer. So that is another exciting area for TCB. So if you look at TCB in general, we have a wide customer base. We have a very diverse applications. We don't just depend on certain applications, but a very diverse application both on the logic side as well as on the HBM side. So certainly, panel packaging at CoP level is an interesting development for us as well.

Also, if I may, I do want to ask a question on SMT. So any update on your strategic review for the division? And have you identified a specific option that you want to go for? And what would be the time line?

To answer your second question first, no. I think we are still in the progress of evaluating. But certainly, we have received some interest in the SMT business at this point in time.

Next, I would like to request Arthur. Arthur, please unmute.

Robin, can you hear me?

Yes.

Congrats on a strong result and guidance. So I will have two questions. Number one is on the CPO. You just mentioned that you have a deep dive into the process, right? And you mentioned the timing is '28. I want to confirm that that's the optical engine shipment or that's your equipment shipment timing?

Shipment TAM, we're referring to our equipment TAM for CPO, it wouldn't look interesting from '28 onwards.

Okay. Interesting on '28. Because what we heard from the supply chains that actually some of the CPO equipment already kick off. So can you share more color on your target for example, is for the GPU side maker or is for the ASIC side maker?

I think the clientele that we have probably serving both Arthur. I think can really differentiate whether it's ASIC or GPU, yes.

Okay. Okay. And my second question is about the HBM. So thanks for sharing this good progress. Do you think for the HBM4 and 4E and actually, you can continue to -- so my question is about HB progress, hybrid bonding progress and your tools, so do you think the visibility is getting longer and longer in the HBM side?

Not really. I think the road map from our customers are pretty clear, but like what Katie said, if there is adjustment to the road map, of course, the demand will vary from quarter-to-quarter. But however, in the long run, I think we are still sticking to a very significant TAM of $1.6 billion in 2028. And we intend and we have never wavered from our aim to target 35% to 40% market share for the whole TCB market.

Got you. And finally, a question to Katie. On the modeling perspective, we know there is a strong booking, right? But how about your component supplier lead time? Is it getting longer? Or does that remain controllable? How should we think of the real billing seasonality of this year? Do we see any constraints?

Yes. Arthur, thank you for the question. Maybe I'll answer it by segment. On the SMT front I'll do that first, though we have very, very strong bookings, as we've mentioned a few times now, the conversion -- the revenue conversion is somewhat impacted by the lead time of our suppliers. The team is actually actively addressing this, and we expect that in the second half, this kind of situation actually will get better.

On the SEMI side, I mean, there's always the kind of tight supply chain, especially given all the uncertainties around the globe. But so far, we will say that we are managing the supply situation just fine.

Next, I would like to request Alex Chan to unmute. I think Alex have some technical problem. Maybe we'll go to the next one. Next, I would like to invite Donnie. Donnie, could you please unmute?

Can you hear me?

Yes, Donnie. Go ahead.

Robin, my first question is regarding to your NEXX business. So I'm wondering if that -- I mean, I remember in the past few years, the NEXX one of the NEXX major business was plating tool, and it can be sold to some PCB companies. And recently, I think PCB companies or substrate companies are expanding capacity due to their running out of the fab. So I think I just want to ask is like if NEXX can generate some revenue momentum recovery in the future? Are you still considering to sell this business? And also in terms of time line, when we expect that we can dispose SIPLACE and NEXX, these two businesses in the future? Would that be in second quarter? So this is the first question.

I think let me answer your second part of the question first. We don't have an exact timing for you. Whenever there is a deal we have we will announce it. But so far, we have nothing to mention here. Now you're talking about NEXX, yes, they are into deposition. When we make a decision to divest NEXX, we don't just look at financial alone. We look at strategic fit to our whole business, right? So we feel that we want to divest that because we're going to focus more on the SEMI back-end. NEXX is not exactly in the back end. NEXX is on the middle end. So that's the reason why we made the decision to divest NEXX, not purely on financial but because of strategy.

Okay. Okay. Understood. Just one follow-up on this. So when you decide to divest NEXX, have you already seen the pickup of the orders from those PCB and substrate makers?

I would say, yes, as I said, but our decision is not based on financial alone. It's really more on strategy.

Okay. And my second question is regarding to the CPO. So my understanding was that AMICRA for example, can be used for laser bonding or, as you s, maybe micro lens bonding on to IC. I guess that's the major business still today. But I'm also wondering if you can quantify a bit more on the 5x growth in the first quarter of this year. It's like what kind of base in the last year? And also in terms of the inspection, as you know, for CPO, the FAU alignment with optical engine is also very critical and it also requires inspection. So we have AOI tool. We have AOI tool. I'm wondering if we can explore some of the business opportunities there or we are mainly staying at the die bonding market?

Yes, very good question, Donnie. Actually, when we deep dive into the Photonics business, right? So we also think that may be we shouldn't just focus on just on the die bonding because there's indeed a lot of opportunities in the photonics in the CPO as well as the optical transceiver business. So too early. I don't have any concrete answer for you at this point in time. But coming back on the FAU, yes, I think AMICRA has a solution especially for FAU attached onto the PIC. So as I said earlier, I think as far as CPO is concerned, we feel good. We feel excited about this particular development. So we will probably have more of the share as we move throughout the quarters in the years to come.

And in terms of 5x growth, can you elaborate a little more is like what kind of base we are growing from in

capacity.

Yes. I would say Donnie, right now, it's still a small base. But again, in terms of growth, if it was a significant growth we thought it's worth to highlight to you guys as well that we are making good progress in terms of optical transceiver as well as CPO.

I'm afraid this is the time that we have. And now I'll pass the time back to Robin for his closing remarks.

Yes. So thank you for a very good discussion today. So let me take a step back and say that this quarter really marks an important point for ASMPT. We delivered one of the strongest quarters in recent years, not just in terms of revenue and bookings, but notably also how broadly AI is translating into more opportunity for ASMPT, from TCB and advanced packaging to photonics, CPO and mainstream platform, we are indeed seeing AI driving demand across multiple products and customer segments at the same time.

Also this breadth matters because it reflects the increasing complexity of AI system architecture and the value of back-end manufacturing, an area where our range of solutions, our scale, our capabilities are allowing us to participate meaningfully across the technology space. We are indeed very encouraged by the operational leverage we have demonstrated this quarter. Our adjusted margins improved sequentially, supported by product mix and volume, and our results came in ahead of market expectations.

So looking ahead, we continue to see AI as a multiyear structural driver of our business, with strong engagement across advanced logic and memory photonics and CPO and mainstream wire and die bonding and SMT pick-and-place solutions. We believe ASMPT is well positioned to support this next phase of industry growth across both SEMI and SMT.

So once again, thank you for your interest and your continuous support. We look forward to updating you more in the next quarter. So this concludes our call. Thank you, and take care.

Good morning, ladies and gentlemen. I'm Ben Poh, Head of Investor Regions and today, I will be moderating the call. On behalf of ASMPT Limited, welcome to our fourth quarter and full year 2025 Investor Conference Call. Thank you all for your interest and continued support. [Operator Instructions]

Before we start, let me go through our disclaimer. Please note that there may be forward-looking statements about the company's business and finances during this call. Such forward-looking statements could involve known and unknown uncertainties, risks and could cause actual results, performance and events to differ materially from those expressed or implied during this conference call. For your reference, the Investor Relations presentation on our recent results is available on our website.

On today's call, we have the group Chief Executive Officer, Mr. Robin Ng and the Group Chief Financial Officer, Ms. Katie Xu. Robin will cover the group's key highlights for the fourth quarter and full year 2025 and provide outlook and guidance for the following quarter. Katie will provide details on the financial performance for the year and quarter.

Now I will hand the time over to our Group Chief Executive Officer, Robin?

Thank you, Ben. Good morning. Good afternoon and good evening, everyone. Thank you for joining us today for our fourth quarter and full year 2025 earnings conference call.

Before we begin, and as I'm sure you know by now, I recently announced my decision to step down from Erol as Group Chief Executive Officer for personal reasons and to devote more time to my family. I will remain in my role until the successor is appointed to ensure a smooth and orderly transition. I'm proud of what we have achieved as a business during my time as CEO and I'm grateful for your trust in me over the years. I'm confident that ASMPT has the right foundations and the people in place for its next phase of growth. Thank you once again for your continued support.

Moving on. The group has decided to divest ASMPT nacks, which has been classified as a discontinued operation. Therefore, -- please note that unless otherwise specified on today's call, we will refer to the group's continuing operations only.

Now for the key highlights for 2025. We experienced strong performance in both our semi and SMT businesses, supported by AI-driven structural growth. There was an increase in customer activity translating into meaningful bookings and revenue for the group, evident in both advanced packaging and our mainstream portfolio. Group bookings grew 21.7% year-on-year driven by both SMT and semi businesses and our full year revenue increased 10% year-on-year, mainly from our flagship TCB solutions.

Now let's look at TCP. TCB momentum strengthened further in 2025 with significant new orders across logic and memory, solidifying our TCB technology leadership. We established deep engagement with both logic and memory customers and saw encouraging traction in areas such as HBM and C2W ultra fine pitch applications. This continues to reinforce our position as a leading provider of advanced packaging solutions as customers move to more complex chiplet-based and high-density architectures.

Turning to our SMT segment. Bookings were better than expected, supported by AI servers, China's EV ecosystem and increased requirements for data transmission for base stations. Last but not least, we also advanced several transformation initiatives from late 2025 to date. These are to enhance focus on our back-end packaging business, improve agility and optimize our portfolio as part of a longer-term strategy. These actions will place us in a stronger position to scale capabilities in the areas where customer demand is more structurally aligned with our technology strength.

Overall, 2025 was a year where we executed well, deepen customer engagements and continue building the foundation for sustained growth. I will elaborate further as we move to today's presentation.

Let me now provide an update on the TCB total addressable market. This time last year, when we presented this slide, we expected the term to reach around USD 1 billion by 2027. Since then, the landscape has evolved meaningfully. The acceleration of AI-driven investment especially in advanced logic and high-bandwidth memory has expanded the market significantly more than our earlier assumptions. Based on our latest projections, we now estimate the TCB TAM to grow from roughly USD 759 million in 2025 to USD 1.6 billion by 2028, representing a CAGR of 30%. This reflects sustained adoption of 2.5D architectures, higher HBM stack and the industries move towards final pitch interconnect. All areas where TCB is increasingly the preferred solution.

Our target market share remains at 35% to 40%. This is supported by the breadth of deep engagements across leading logic and memory customers and by the performance of HBM, C2S and C2W TCB platforms, including strong uptake of our plasma enabled ultra-fine pitch capabilities. We are well positioned to benefit from this expanded TCB TAM, and we are committed to continue investing in this exciting technology.

Moving on to advanced packaging. This remains a strong growth engine for us in 2025 supported by rising complexity in both logic and memory packaging. As customers shift further towards chiplets highest at HPM and final pitch interconnects, we continue to see solid demand across our TCB platforms, in particular. Of note, with our breakthrough into comparative fish beer market, we also grew TCB market share significantly, achieving record TCB revenue growth about 146% year-on-year. In 2025, our AP revenue growth of 30.2% year-on-year was driven by TCP. As a result, AP's contribution to group revenue also increased from 26% in 2024 to 30% in 2025.

Now let's look at TCB more closely. In logic, our C2S solution maintains its dominant position as a process of record with a steady flow of orders from key OSAT customers in 2025. Extending into early 2026, we are pleased to share that we have secured additional orders for 9 more TCB tools from the same customer. We are well positioned for further order wins as the market shift towards larger compound lines. At the same time, our C2W ultra-fine pitch platform, enhanced with plasma EOR technology secured orders for 2 tools in February 2026 from a leading customer for C2W applications. Since the announcement, we have secured 2 more such tools from the same customer. As the industry transitions from master technology to TCB, the group stands to benefit significantly as the preferred C2W solution provider offering plasma enabled capabilities.

This engagement underscore the confidence customers place in the ability to support tighter technical specifications and next-generation packaging road maps. In memory, we deepened our engagement with several customers and continue to expand our share with shipments in Q4 2025. Our tools have demonstrated superior performance with industry-leading production yields and interconnect quality. We were also the first to secure HVM for 12 high orders from multiple players, and we are now leading HVM-4 16 development with flex-based TCBI deployed for sampling, and our fluxes AOR-TCB process under qualification.

These are important milestones for our technology leadership as HBM architectures scale further. Beyond TCB, we also made progress in hybrid bonding, where we receive customer buyout and ship modules. Our second-generation hybrid bonding solution is highly competitive, offering high alignment precision, bonding accuracy, footprint efficiency and units per hour.

In Photonics, revenue grew year-on-year, and we sustained our leading position in the 800G optical transceiver market, while continuing development work with industry partners on 1.60 transceiver solutions. Our CPO collaboration also continues to move forward with key global players. And in SMT SIP applications, Demand remained robust, especially in AI-related RF and system in package application. Our next-generation chip tool also gained traction among advanced logic smartphone applications. Overall, Advanced packaging delivered another year of meaningful progress with broader adoption across logic, memory, photonics and SiP and it continues to be a central pillar of our long-term growth.

And finally, our mainstream business. This accounted for about 70% of fiscal year '25 group revenue. In 2025, AI-related demand was also a strong momentum driver for our mainstream business. Rising requirements for AI data center power management applications, kept utilization reach elevated and leading global IGMs, benefiting semi mainstream. Meanwhile, SMT mainstream secured more orders to support increased data transmission requirements for base stations and AI server [indiscernible].

In China, our mainstream business saw around 18% year-on-year revenue growth across both semi and SMT. Semis growth was driven by strong demand for in and diode applications underpinned by robust set utilization. SMT benefited from increased deployment of AI server bots and strong demand for Ebix in 2025.

With these highlights, let me now hand over the time to Katie, who will walk you through our group and segment financial performance.

Thank you, Robin. Good morning, good evening, everyone. Let me take you through the group financial performance. Before I start, I would like to reiterate that unless otherwise specified, the numbers I will be referring to today are for the group's continuing operations only, with adjustments made under non-HKFRS measures. This slide covers our financial results for 2025.

For the full year, the group delivered revenue of USD 1.76 billion, representing an increase of 10.0% year-on-year, driven largely by TCB. Group bookings reached USD 1.86 billion, representing 21.7% year-on-year growth. Both SMT and semi registered high bookings during the year. The group continues to build a healthy backlog with book-to-bill of 1.05, which is our highest since 2021. In 2025, group adjusted gross margin was 38.3%. This was 172 basis points lower year-on-year, reflecting lower gross margin in both SMT and semi.

Group operating expenditures was HKD 4.56 billion, up 3.2% year-on-year, mainly driven by strategic R&D and IT infrastructure investments of HKD 237 million as we communicated at the beginning of last year. These investments were partially offset by disciplined execution of cost control and efficiency measures.

Now looking ahead for 2026 packs, as Robin mentioned, we are committed to continuing the investment in our core technologies, and we expect OpEx to rise by about HKD 200 million in 2026. In 2025, both adjusted operating profit and net profit improved year-on-year due to high revenue and operating leverage. In the fourth quarter, we delivered revenue for continuing operations and discontinued operations of USD 557.1 million that surpassed the upper end of our guidance. Q4 revenue for continuing operations was USD 508.9 million, representing an increase of 12.2% Q-on-Q and 30.9% year-on-year, driven by strong growth across both semi and SMT. Group Q4 bookings were USD 499.7 million. The Q-on-Q increase was due to stronger TCB bookings, while the year-on-year growth was largely driven by SMT's mainstream business.

Group Q4 adjusted gross margin was 35.8% and down 175 basis points Q-on-Q and 101 basis points year-on-year. This sequential decline came from both semi and SMT with year-on-year decline due to lower semi margins partially offset by higher SMT margins. Group Q4 adjusted operating profit was HKD 161.0 million, up 4.3% year-on-year due to -- up 4.3% Q-on-Q due to higher revenue and operating leverage. Group Q4 adjusted net profit was HKD 119.9 million, up 42.2% Q-on-Q and 390.7% year-on-year. The Q-on-Q increase was largely due to fees of HKD 39 million from order cancellations while the year-on-year increase was due to stronger operating profit. Adjusted earnings per share were $0.30.

Moving on to the Semiconductor Solutions segment for the fourth quarter of 2025. We Semi delivered Q4 revenue of USD 245.6 million, an increase of 9.4% Q-on-Q and 19.5% year-on-year. Q-on-Q and year-on-year growth was driven by AI-related applications, mainly from Photonics. Semi Q4 bookings were USD 253.3 million, up 15.4% Q-on-Q and 2.3% year-on-year. The increases were due to TCB orders from advanced logic customers and a market share gain in high-end die bonders. Semi book-to-bill ratio in Q4 2025 was $1.03. Q4 adjusted margin for semi came in at 40.3%, down 102 basis points Q-on-Q and 292 basis points year-on-year. The Q-on-Q decline was largely due to product mix and inventory provision as a result of an isolated order cancellation.

Year-on-year decline was due to product mix, inventory provision mentioned above and higher factory utilization in Q4 2024 during the TCB ramp. Q4 adjusted segment profit was HKD 98.0 million, up 62.5% Q-on-Q and up significantly year-on-year. Both Q-on-Q and year-on-year improvements were mainly driven by higher volume and fees related to the other cancellations.

Next, let me move to the SMT Solutions segment performance for the fourth quarter of 2025. SMT delivered strong Q4 revenue of USD 263.3 million, up 15.0% Q-on-Q and 43.8% year-on-year driven by AI servers, EVs in China and the billing of a bulk order for smartphone applications. However, contributions from automotive end market outside of China and industrial remains soft. SMT recorded Q4 bookings of USD 246.4 million, down 3.9% Q-on-Q but up 73.3% year-on-year. The Q-on-Q decline was due to seasonality, while the year-on-year increase came from the demand for AI servers and EVs in China. Q4 SMT gross margin was 31.6%, down 225 basis points Q-on-Q, but up 199 basis points year-on-year. The Q-on-Q decline reflected continued weakness in automotive and industrial end markets and the building of bulk order mentioned above, which had a lower margin. The year-on-year increase was mainly due to higher volume.

Q4 segment profit was HKD 193.1 million up 18.5% Q-on-Q and significantly year-on-year due to higher volume.

This slide highlights ASMPT's revenue breakdown by end markets. Computer end market was significantly up, becoming the largest contributor to group revenue, accounting for 22%. The growth in computing was largely driven by our TCB solutions. Consumer end market was the second largest contributor at 17%. Year-on-year revenue growth came largely from the group's mainstream solutions, consistent with higher revenue from China. The communication end market contributed to 16% to group revenue, driven by photonics and high-end smartphone-related applications. The automotive end market contributed almost 16% to group revenue, supported by EV demand in China, where the group remains a leading player. Lastly, the industrial end market contributed 10% to group revenue, reflecting soft market conditions.

As you can see from this slide, we're a truly global business partnering with customers across all major regions. China remained the largest market, contributing 41% of group revenues. However, Europe and Americas declined year-on-year, mainly due to soft market conditions in SMT with Europe's share of revenue down to 13% and Americas down to 11%. Looking at Asia outside China, their proportion increased collectively from 24% to 34%, largely driven by TCB revenue. The group continued to maintain low customer concentration risk with the top 5 customers representing approximately 16% of total revenue in 2025. We have an existing dividend policy of disturbing about 50% of the annual profit as dividends, and we firmly believe in returning excess cash to our shareholders. For the second half of 2025 with adjusted EPS at $0.68 in Hong Kong dollars, while continuing and discontinued operations, the Board has recommended a final dividend of $0.34 per share. In addition, the Board has recommended a special cash dividend of $0.79 per share after taking into consideration the net cash inflow from recent strategic projects. Together with the interim dividend of $0.26 per share paid in August 2025, the total dividend payment for 2025 will be USD 1.39 per share.

With that, let me now pass the time back to Robin for an update on our transformation initiatives and the next quarter's revenue guidance.

Thank you, Katie. As mentioned earlier, we undertook several transformation initiatives from the late 2025 to date as part of our long-term strategy. In November 2025, we completed the divestment of our entire equity interest in AMI in exchange for cash and new shares in Santen original advanced compounds Company Limited. In January this year, we announced a strategic options assessment of our SMT Solutions segment. The assessment is underway, and we will update at the appropriate time when there are material developments.

Lastly, today, we make public the decision to divest ASX incorporated. These initiatives share a common objective of optimizing ASMPT's portfolio streamlining operations to enhance agility and improving margin and profitability while ensuring continued investment in infrastructure and technology development in high-growth areas. We also sharpened our focus on the back-end packaging business. In the meantime, business for all our segments continue as usual.

Let me now turn to our Q1 2026 revenue guidance. The group expects Q1 2026 revenue to be in the range of USD 470 million and USD 530 million. At midpoint, this represents a decline of 1.8% Q-on-Q and 29.5% year-on-year. Notably, the group's midpoint revenue guidance for continuing operations only already excites talent market consensus, which includes both continued and discontinuing operations. We anticipate sustained Q-on-Q revenue growth in our semi segment driven by TCB and high-end die bonders, although this will be partially offset by SMT seasonality. On a year-on-year basis, the higher group revenue is expected to be driven mainly by strong momentum in SMT coupled with steady growth of semi.

For Q1 2026, group gross margin is expected to improve, led by semi gross margin returning to the mid-40s level. This improvement is driven by higher volumes from TCB and high-end die bonders. SMT's gross margin, however, is expected to stay at similar levels as automotive and industrial end markets remain soft. The group bookings momentum will accelerate in Q1 2026, supported by both segments.

Looking further ahead, structural industry growth from AI demand is expected to drive revenue growth across both semi and SMT. In TCB, with our industry-leading technologies, and deep engagement across a broad AI customer base, we are well positioned to expand our TCB business in a rapidly growing market. Our semi and SMT mainstream businesses continue to be supported by global investment in AI infrastructure and steady demand from China, while SMT automotive and industrial end markets are expected to remain soft in the near term.

This concludes our full year and fourth quarter 2025 presentation. Thank you, and we are now ready for Q&A. Let me pass back the time to Ben to facilitate.

Thank you, Robin. Ladies and gentlemen, we will now begin the Q&A session. [Operator Instructions] So with that, may I have the first question. Okay. Gokul, please omit yourself and raise your question.

Ben, Robin and Katie, Robin, first of all, thanks for your leadership over the many years, and good luck on your retirement. My first question is on TCB, the addressable market TAM expansion to $1.6 billion. Could you talk a little bit more about where is the upside mostly coming from in your estimates? Let's say we get to this $1.6 billion, what will be the mix of HBM versus logic look like in 2028? And given that you gave an estimate of $750 million addressable market for last year, what was the market share roughly for ASMPT last year? Should we assume that it was about 30% or so for TCB, just to get a starting point of your TCB journey when we think about this TAM expansion?

Gokul, this is Katie. Let me try and address the questions that you have. First, on the TCB TAM, just take a quick minute on the methodology. Actually, last year, that was our first time publishing the TAM at $1 billion. This year, actually, the methodology is very, very similar. So we essentially used the wafer per month that actually you guys have published in the industry, and we start that to the number of AI chips and the interconnects and then choose needed, right? So it's the same methodology. So to your question about what's driving the expansion? Obviously, right, really, it's the starting point is the wafer per month that has expanded significantly for the AI industry overall. So that's the main expansion.

Now, in terms of the mix of hyper bond -- sorry, HBM and logic, I think in previous years, we've communicated HBM is the larger portion of the TAM. And it'll continue to be so probably until as we go into the outer years, right, if we talk about HBM 20 high beyond, then at that point, maybe hype-bound will be kicking in and then the logic side, especially COW, will actually become more prominent in the TAM.

So then the other things you asked about the last year's market share and you said about 30% and you are quite in the ballpark for that one.

Got it. That's very clear. Second, on the proceeds from, I think, this rationalization of the portfolio and some strategic actions that you're taking. Good to see that happen. But could you also talk a little bit about what is the kind of end state that you're hoping for once this rationalization is being done? Are there areas that you're kind of trying to bulk up on as it putting to the back-end packaging business? And specifically on NEXX, what is the rationale for divesting NEXX given that has a fair bit of 2.5D bumping and ECD plating kind of business, which theoretically, it feels like closer to the advanced packaging business. But just help us understand why that divestment of NEXX is also happening.

Gokul, thanks for the question. I'll take that question, Gokul. So basically, I think it's really focusing really on our back-end packaging business because this is where we feel this is where the structural growth will be -- and this is where I think our strength sort of match the industrial road map for packaging. So really back to focusing on back-end packaging. So first, you notice we divest our late crane business just as just 1 step. And now we are assessing SMT, which is more the downstream operation.

And then as to your question on NEXX, you're right. NEXX is -- although it is advanced packaging, but it's not exactly back end. -- belong more to the middle and -- and their technology, to be honest, is more wet technology, whereas technology is really more on automation, innovation and so forth. So we feel that it's probably the right time to consider divesting mix next to really focus all our attention and all our resources on the back end side.

Understood. And maybe if I could squeeze in 1 more. I think any quick view on how the mainstream DME solution business you're expecting it to progress, Robin? What are you hearing from your customers given at least from a CapEx perspective, many of your customers seem to be moving up for the first time in this up cycle?

Yes, yes. I think we're beginning to see maybe we talked about green shows some quarters back. But this time around, the ratio seems to be real from our point of view. Now because there is a tier wind behind the mainstream business. And this time around, we feel that -- we have been talking for a few quarters already, Gokul, that we feel this standalone is underpinned by AI investment as well. You can imagine when the industry continue to invest more and more in terms of data center besides the GPUs, there are many other components inside the data in the sites the servable AI pool. You have power management devices and many other components, right? So you can imagine with all the server bots going into a data center and the build-out of data center CapEx. There is huge massive amount of components need to be packaged using both our semi wire bond and the normal die bond tools as well as our SMT pick-in-place tools. So this AI data center investments are really driving our mainstream, both on the semi side as well as on the SMT side.

Okay. Okay. That's very clear. So we should expect that mainstream semis also should be growing. I think it's not been growing for maybe 3, 4 years now after 2021, but it looks like '26, we should see some growth in the non-advanced packaging piece of semi solutions as well, right?

Yes. As far as you can see, I think our visibility is, again, is quite normal in our business to be limited to 1 or 2 quarters, right? So I think first half looks to be okay. Half-on-half better than -- half-on-half growth year-on-year, half year also, we think it will grow. But if you ask me on the second half, let's wait for a while to see how it develop limited visibility at this point in time for second half.

Thank you, Gogo, for your questions. SP1 I see a raise hand from Daisy. I will request Daisy to unmute and raise your question.

Firstly, I want to ask about the HBM opportunities because I listened to your competitor's earnings call, they are talking about high-bandwidth flash opportunities. Have you guys also seen these opportunities from ASMPT side?

Yes, we do, Daisy. This is a very good question. I think this is, again, probably an exciting development. To be honest, we have not factored this into a TAM, TCB TAM because potentially, the way we assess the technology or the packaging technology required, I think TCB could be also be a tool to package HBF. So this is something that we look forward to. If the industry -- if the industry develop in this direction, I think we will also stand to benefit in time to come.

Okay. And also following Gokul's previous question and you previously also mentioned that you expect the second half will also grow versus first half. So I want to ask about the order visibility for -- from ASMPT side because I think in normal times, back-end order visibility is 3 to 6 months. And how is the order visibility now? And what is the magnitude that you are seeing that second half could grow versus first half?

Correction, correction, Daisy correction. Let me be clear. Just want to answer Gokul's question. I'm just saying first half 2026, we have better visibility because of the momentum we are seeing in terms of advanced packaging as well as mainstream. But second half is still limited in terms of visibility. But at least this time around, we can see a little bit further, maybe slightly more than a quarter, but second half, let me correct your statement, second half we still have limited visibility. So when I mentioned just half-on-half I'm sort of giving you some color. First half this year, demand probably will be better than first half last year as well as second half of last year. So I'm just comparing half-on-half and year-on-year. But second half, I repeat, we still have limited visibility at this point in time.

Okay. Thank you, Daisy. And next, request [indiscernible] to unmute.

Robin, you will be missed. First, congrats on the strong results. So first question is on the backlog. You highlighted that backlog almost over $800 million. Can you give us more color on the spread between the semi and SMT? And also, you highlight the high-end bonder. Can you share with more on the TCB's product such as panel label fan-out?

[indiscernible], on the backlog, just really quick. The semi signed backlog is stronger as a large quantum than SMT.

Is this a significant higher or is pretty Estima? Yes.

Met the percentages roughly 60-40, I guess, don't call me that exact, somewhere there.

So [indiscernible], on your second question about high-end Dion, which I think you're referring to what we have mentioned in our announcement, Yes, I think if we're referring to semi, I think it's good news. We have penetrated into a high-end die-attach application for high-end smartphones, right? So if you look at the camera modules or high-end smartphones, there are many, many box ship components in there, which need to be put in place as well. So the customers have chosen our die-attach application to place those components. So this is a brand-new market for us. We have never been in this market. So we really look forward to having more market -- increasing the market share in this particular area. So that's for the high-end in bond I think you're referring to.

Now you also have a question on panel-level [indiscernible]. We see a lot of trend in that direction. We feel that this is also driven by AI as well, right? So panel level [indiscernible] for components that go into their center becoming more and more visible. So definitely, we have a tool, basically a mass reflow tool that we can deploy for a solution that is so we're also bringing well placed to capture this opportunity.

Got you. And second question is on Page 10. You highlight there is older cancellation on the semi side. Can you give us more color? Is it associated with the NEXX?

Yes, the -- so this audit cancellation does not have an association with NEXX. So let me just give a little bit more color on that. The order cancellation came from a global IDM who's focused on automotive applications and other came a few years ago and it was for our semi mainstream products. The customer had to cancel the order due to weak automotive industry performance. So that's why we got this cancellation. But I want to make sure that well, understand this is a very much an isolated event.

Next, I would like to request Leping to unmute and raise the question.

The first question is also about the TCB TAM. So when you derive the TCV time in 2028? What's the split between memory and logic? And you also said that you're targeting 35% to 40% market share in 2028. So what's your current market share in memory and logic and what's the upside we can expect in the next few years?

Leping, maybe just add a little bit more basically essentially the answer provided a on the split of memory and logic. Currently, the memory -- the HBM portion in the TAM definitely is much larger than logic. But as we go out -- a few years out, this dynamic will actually shift where the larger -- especially COW should take a larger share. But we do not share the specific split for confidentiality reasons or competitive reasons, I should say.

Now in terms of the market share, as Robin has mentioned in the opening, ASMPT is very, very strong in COS and also when we are actually making all of wins on COW. So our market presence in the logic space is very strong. On HBM, you guys probably remember a year ago, we broke into HBM market. So we have gained market share there. So that's kind of where we are in terms of market share.

Maybe just to add on a little bit in terms of the competition landscape, I think in the logic space, we are a for a very key supply chain for iron use application. And then recently, the good news is that we announced we won 2 tools for C2 fuel application, right, for the same supply chain, and then we won 2 more. So I think it is a signal that we are also being recognized as a solid solution provider for the CDW space as well. Now on the memory side, I think the competition landscape is different. We have a strong company in the memory space. But we have done a fantastic job in 2024. We have practically 0 share in HPM. And in 2025, we managed to penetrate in a very meaningful way. in the HBM market. Now that we are -- we have a strong foothold in the memory market. We look forward to better times ahead in terms of HBM demand allocation.

Okay. The second question is about the memory super cycle. So are you seeing an acceleration of the capacity expansion from your HBM customer -- and given your HBM 4 of order win, are you customers provide a longer-term rolling forecast to secure your TCB tool for the 12 high and high? Or how you play your capacity in this year for the TCB business?

Yes. Thanks. Definitely, definitely in terms of the ABM CapEx is really in line with investment in data center, right? So data center investment continue to increase. You can expect HBM to continue to increase as well, not just in the number of HBM but also in the highest stack from 125 to 165 to 20 high potentially. So that means there will be more and more opportunities for TCB packaging as HBM continue to stack up in terms of high.

Now you asked whether about capacity allocation, to be honest, I think there are some more differentiation between 12 and 16, but they are not major, some hardware module need to be different. If we use to package 12 between 12 and 16, there are some hardware modifications, but also some software. So not -- so there's not much material differences between the tool in the 162.

Thank you, Leping. And next, I would like to request Simon Woo to unmute and raise your question.

Robin, as always, we'll miss you. So the is long-term pressures for 2028 you are expecting CV market $1.6 billion for 2028. Any rough idea of the percentage of the hybrid bonding assumption for that time or very low single digit to a single [indiscernible]?

Sorry, Simon, because your line is breaking up. So do you mind to say that again?

So my question is that the hybrid bonding portion or the 2028 TAM, $1.6 billion.

So this is the TCB temp, so there is no hybrid bond in the TCB TAM. But I guess you're asking our assumption of the hybrid bond adoption timing. Is that what your question is?

Yes, sure, yes. Yes.

Okay. In our model so far, for HBM 16 high, we assume that and we're actually confident that the TCB will continue to serve 16 high because as we get into 20 high, it really depends on the JTC standard, right? If the standard continues to relax, then there will actually be an upside to this model. Otherwise, we assume that in the model itself that the 20 high will be moving on to Hybean cost.

Partially.

Expect PCB can be used for [indiscernible]?

Yes, Simon. I think looking at how -- looking at the -- how the technology -- CP technology developed over the years and also into the future, we are confident that the TCB technology to get with -- of course, we need to collaborate with our customers as well. They are wafer technology will probably, we believe, will continue to improve. So I think a combination of both the wafer as well as TCB tools, we are hopeful and optimistic that 20 high transdue TCB. Of course, if what Katie said, if the JD standard can be relaxed to increase the high from 75 to beyond 25, maybe 50 or even 1050 micron, then the chance of using more for '20 and beyond will even be higher. So this is a mix. So we just have to wait for a little longer to see how the industry plays out in terms of the high restriction.

Yes, very clear, sir. Do you believe the load earlier core-out PLT will require hybrid bonding at or maybe year or later?

Sorry, you're breaking up. Sorry, I need to ask you to repeat.

Taking on I should use a better one. But my question is alluded earlier -- do you see any meaningful progress for the hybrid bonding for Coast area?

So I believe your question is in the logic area whether there's more opportunity for having bonding, right?

Yes. Correct. Correct.

Yes. Actually, to be honest, hybrid bonding has already been adopted at the chiplet level, right, for certain devices. -- but certainly, but we believe that, that has already been ongoing. It all depends it's very dynamic, right? So even, to be honest, even at a chiplet level TCV can be used as a 2 as well, especially when we look at the exciting technology that we're going to develop for TCB going into the future. The TCB technology will get closer and closer to the hybrid bonding technology. So from that perspective, we are optimistic and hopeful that at some point, TCB can also be used also at a chip integration level. But as I said, this industry is very dynamic. So nobody knows what's going to happen, but let's continue to monitor this space.

Yes. Very clear. Sorry, the last. 30% of your revenues or the bank packaging -- any rough idea -- that means anyway, $0.5 billion of your revenue for the expectation last year. But any other idea what was the TCB potion out of the oven packaging revenue last year?

You're talking about TCB proportion to advance packaging. Is that what I'm saying? .

Yes, 2025. yes.

Very dominant a major share of the AP revenue for TCB.

[indiscernible], majority portion?

Yes, ballpark around it.

If you look at the TCB market, the 10 slides that we shared, and Robin mentioned you know what the TCB market size was in 2025. And I think Gokul mentioned about our market share as you do with the rough calculation, you actually get there, if that's what you guys are trying to do?

Yes, $200 million, $300 million, maybe. Sorry, 1 last question from some investors in what overall the revenue appearance or iron per your metal restructuring for the SMT or lead frame, the back-end area of what percentage of the revenue will be off once you complete the restructuring process?

Maybe let me try to answer your question. So for a clarification. For AMI, we were shareholding. And now after the disposal of AMI, there's actually no revenue impact year-on-year of the last few years. So there's no revenue impact at all. For the NEXX business, we just announced today to be as discontinued business or put up for sale, right? NEXX revenue is about USD 100 million, that's what you're looking for.

Thank you, Simon. Yes, I think we have time for 1 final question. And Donnie, we request you to unmute and raise your question.

Wish Robin you all the best after the retirement -- my first question is regarding to your guidance. Can you break down or elaborate more on the bookings momentum in the first quarter? -- and particularly TCB because I think based on your announcement in fourth quarter last year, we already have received quite some TCB orders. So I'm also wondering what kind of trend in terms of the TCB bookings into the first quarter this year? This is my first question.

Thank you, Donnie. I think you probably expect my answer, we cannot be too granular because for competitive reason as well. But I think in overall, I think 2026, we're expecting TCB to continue to grow in line with the investment -- so much investment in data center, right? So that tens -- now if I drill down to the booking, I'll give you some booking color for Q1 2026. We're likely to see a strong booking in Q1 -- Q-on-Q around 20% growth Q-on-Q and even stronger around 40% year-on-year growth for Q1 booking '26 for both SMT segment as well as the semi segment. I think we have been talking a fair bit over the last couple of quarters as well that we see AP will continue to grow. And because of mainstream momentum gaining very strongly over the last 1 or 2 quarters and into Q1 2026 as well. So I think both advanced packaging as well as mainstream will continue to do well in Q1 2026 as far as bookings are concerned.

Now however, I have a caveat just now as well, right, with the stronger booking -- also let me share bit or qualify that we might see some impact on revenue conversion because we are seeing longer material lead time due to tightness in the supply chain. -- right? So although bookings are going to be very strong in Q1, but the conversion to revenue may take a little bit longer than usual because of supply chain tightness, okay? Yes. So I think this is some color I want to give you. And by the way, I think Q1 bookings, the way we see will be the highest quarterly booking in 4 years.

Understood. Can I have a follow-up on this. So for the semi business bookings, the strong sequential growth -- can we say it's primarily driven by more like conventional packaging or from advanced packaging.

Mainstream will probably grow a little bit more than advanced packaging. Advanced packaging tend to be a bit lumpy. We have been saying that for a long time really don't expect AP revenue to be continuously high because first and foremost customers are limited, less customers than the mainstream second, is a high value tools, so customers cannot continue to buy the quarter-on-quarter. So -- but the demand for TCP is steady for sure, right? But Don't expect this to continue to be on a quarter-on-quarter basis continue to grow. So that's a site. But on -- but what we are seeing quite interesting is really on the mainstream side, -- so we see really a pickup in terms of mainstream for those reasons I said earlier, they are driven data center.

Okay. Got it. And my second question is regarding to your 2025 review in terms of the market share gain, particularly in the HBM market. So -- but if you -- if I remember correctly, we actually received quite sizable orders from fourth quarter 2024 from leading HP customers -- and since then into 2025, actual bookings despite of -- there are some repeat orders, but it seems like not as significant as what we had back in fourth quarter 2024. So I just want to clarify that our market share gain in 2025 for HBM and TCB is primarily driven by the big orders we received in fourth quarter 2024?

Just really quick. Donnie, the market share data is actually based on billing.

Yes. So the follow-up is like -- when should we expect to receive a more meaningful repeat orders from the leading HPM customer? I mean -- or when should we can be expect that orders can be maybe more significant than what we had back in fourth quarter 2024?

Yes. I think it all depends on how soon the rollout in volume for section high, right? So also that depends on their customers rollout of the new architecture. So the timing has to be aligned with ultimately how all the ultimate consumer rollout the GPU architecture. So I think as the industry moved from 12.5 to 16 high, I think all equipment supply, including yourself for TCB are waiting anxiously for that particular customer to allocate TCP demand. So -- so at this moment, we feel that 2026 will be a year whereby there will be new to it for [indiscernible]. But exact timing, unfortunately, Donnie, I cannot give you any visibility at this point in time. But it cannot be too long basically no opinion.

Thank you, Donnie. That will be our last question for today. So I will pass the time back to Robin for his closing remarks.

So thank you for all your well wishes, but my retirement. Now before we end, let me capture some really key takeaway from today's discussion. First, 2025 was a year of solid execution for us. and strengthening our customer engagement across the group. So we delivered growth in both bookings and revenues with a book-to-bill ratio of 1.05 and a healthy backlog, reflecting continued momentum and trust the customer placing us Second, AI-related demand was the engine of our overall business in 2025 across both infrastructure and applications, AI drove significant activity in both semi and SMT. This reflects an enduring structural trend that we expect to persist for some time as we increasingly shift customer road maps and priorities.

Last but not least, TCB was a standout for us in terms of momentum and in terms of technology leadership. We expanded envision in both logic and memory, securing wins across HBM, C2S and CW application. So with our latest CCB 10 projection, this highlights the scale of the opportunities in TCB, and we continue to target a 35% to 40% share of this market. So in short, before I close, overall, we are well positioned as we enter 2026. So thank you once again for joining us, and we look forward to updating you in the next quarter. This concludes our call. Thank you, and take care.

Good morning. Ladies and gentlemen, this is Ben Poh, the Head of Investor Relations at ASMPT. And today, I'll be moderating the call for the first time. On behalf of ASMPT Limited, welcome to our third quarter 2025 investor conference call. Thank you all for your interest and continued support. [Operator Instructions] During the Q&A session priority will be given to the covering analyst.

Before we start, let me go through our disclaimer. Please note that there may be forward-looking statements about the company's business and finances during this call. Such forward-looking statements could involve known and unknown uncertainties and risks that could cause actual results, performance and events to differ materially from those expressed or implied during this conference call.

On the call, unless stated otherwise, all references to gross profit or margin, operating profit, segment profit and net profit are on adjusted basis as described in our MD&A. For your reference, the Investor Relations presentation on our recent results is available on our website.

On today's call, we have the Group Chief Executive Officer, Mr. Robin Ng; and the Group Chief Financial Officer, Ms. Katie Xu. Robin will cover the group's key highlights for the third quarter, guidance and outlook for the next quarter, while Katie will provide details on the financial performance for the third quarter.

Now I will hand it over to our Group Chief Executive Officer, Robin.

Thank you, Benjamin. Good morning and good evening to everyone today. It is a pleasure to have you all on our earnings conference call for the third quarter of 2025. Now let's start with the key highlights of the third quarter.

This quarter, we continue to experience strong momentum driven by AI. The group's Advanced Packaging and mainstream businesses continued to benefit from sustained AI adoption. The group's strong Advanced Packaging momentum has been driven by Thermo-Compression Bonding or TCB. We remain dominant in advanced logic, have made rapid inroads into high-bandwidth memory or HBM and more recently have a first mover advantage in HBM4.

At the same time, AI infrastructure comprising data centers, data transmission and power management contributed to demand in the mainstream business. In China, demand was also driven by EV and high factory utilization across OSATs. Now let me talk about our technology leadership in TCB.

We have further solidified our leadership in HBM. The group's HBM TCB solution have achieved better yields versus the competition. And as I said above, we are leading in the transition to HBM4. In addition, our proprietary fluxless active oxide removal technology provides superior scalability for HBM 16-high and above with the lowest cost of transition.

In logic, the group's ultrafine pitch TCB for chip-to-wafer with plasma AOR solution has successfully passed final qualifications for quality and reliability at a leading foundry and is ready for high-volume manufacturing. Notably, plasma-based technology has been endorsed by this leading foundry, underscoring this technological advantage over other processes.

Turning to TCB orders. Encouragingly, the group achieved recurring orders from both memory and logic customers in the third quarter. In memory, our TCB solutions for HBM4 12-high became the first to secure orders from multiple HBM players. We expect to remain as a primary supplier, demonstrating our technology leadership in the rapid transition to HBM4.

In logic, the group continued to win orders as a process of record for chip-to-substrate applications for key customers. As the market transition to a larger compound dies, we are well positioned to secure sizable orders in Q4 2025 and beyond from the OSAT partners of a leading foundry. As a business, we remain confident in the outlook for TCB demand.

As to the other updates, in hybrid bonding, the group continued to ship hybrid bonding tools in Q3 2025. Our second-generation hybrid bonding solutions are competitive in alignment precision, bonding accuracy, footprint efficiency and units per hour. In photonics, we continue to dominate the optical transceiver market, reinforcing our leadership as a key supplier of 800G transceivers while also actively engaging industry players on next-generation 1.6T photonics solutions.

Moving to SMT. Bookings were better than expected in the third quarter, demonstrating signs of recovery in the business. SMT's AP solutions achieved strong bookings year-on-year growth in the third quarter and won sizable Systems-in-Package orders from IDMs and OSATs for RF modules for base station to support AI growth. SMT also continued to win orders for the next-generation chip [ SMT 2 ] in advanced logic smartphone applications from a leading foundry and OSAT partners -- and OSAT players. In our mainstream SMT business, the demand came mainly from EVs where we remain a leading player in China.

Before I conclude this section, I want to highlight that we have delivered a profitable quarter, excluding the strategic restructuring costs from the voluntary liquidation of the Shenzhen AEC plant as announced in August. The decision was made to optimize the group's global supply chain to better align with the evolving market dynamics and customer needs.

As said in the announcement, this move is expected to improve the cost competitiveness, agility and resilience of the group's global manufacturing operation for its key products and solutions.

With those highlights, let me now pass over the time to Katie, who will talk about our group and segment performance.

Thank you, Robin. Good morning, and good evening, everyone. Let me take you through the group financials. This slide covers the group's key financial metrics for the third quarter of 2025.

The group delivered revenue of USD 468.0 million, representing an increase of 7.6% quarter-on-quarter and 9.5% year-on-year, largely driven by growth in SMT. In third quarter, the group recorded bookings of USD 462.5 million, driven by AI momentum. We recorded recurring TCV orders in memory and logic and SMT bookings were also better than expected. This marks the sixth consecutive quarter that we have achieved year-on-year growth.

The group had an isolated bookings cancellation in the third quarter for its panel deposition tools from a leading high-density substrate manufacturer in response to a slower-than-expected digestion of its existing capacity. This is a one-off occurrence. And excluding this cancellation, the group's bookings in the third quarter would have been USD 486.6 million, 1.5% higher quarter-on-quarter and 20.1% higher year-on-year.

The group achieved a book-to-bill ratio of 1.04 for the quarter, maintaining a ratio above 1 since Q1 2025. SMT posted a robust ratio of 1.12, while SEMI's ratio was at 0.96. The group closed the quarter with a backlog of USD 867.7 million.

The group's adjusted gross margin for the third quarter was 37.7%, which is lower than our typical level. It was impacted by a larger contribution from SMT and the lower SEMI gross margin, which I will explain in the next slide. I would like to note that the group's year-to-date adjusted gross margin remained healthy at approximately 40%.

The group's operating expenses were up 6.2% Q-on-Q and 5.3% year-on-year. As expected, higher OpEx was largely due to strategic R&D and infrastructure investments and foreign exchange impact. They were partially offset by prudent spending control and some benefits from restructuring.

The group's adjusted operating profit was HKD 124.4 million, down 26.6% quarter-on-quarter and 30.3% year-on-year due to lower gross margin and higher operating expenses. Group adjusted net profit was HKD 101.9 million, down 24.4% quarter-on-quarter, but up 245.2% year-on-year.

The quarter-on-quarter adjusted net profit, which included the fee collected from the order cancellation mentioned above was offset by the absence of tax credits recorded in the previous quarter. The year-on-year increase in adjusted net profit was driven by the fee collected from the order cancellation and a lesser negative impact from foreign exchange. The adjusted earnings per share was HKD 0.24.

Now moving on to the Semiconductor Solutions segment for the third quarter of 2025. SEMI's revenue was USD 240.5 million, down 6.5% quarter-on-quarter, but up 5.0% year-on-year. The year-on-year revenue increase was driven by stronger demand for wire bonders and die bonders due to the increased needs for power management across multiple applications. Quarter-on-quarter revenue decline was due to the timing of key customers' AI technology road maps, which impacted AP demand this quarter. There was also some shipment disruption caused by a typhoon in September in China.

SEMI's bookings of USD 207.8 million were down by 1.7% quarter-on-quarter and 12.4% year-on-year. Excluding the booking cancellation explained above, SEMI's Q3 2025 bookings would have been USD 231.9 million, 9.6% higher quarter-on-quarter and slightly lower year-on-year. SEMI recorded quarter-on-quarter and year-on-year growth in wire bonders and die bonders. TCB orders were up quarter-on-quarter but remained at a lower level due to the impact on AP demand, as mentioned above.

As I said earlier, SEMI's adjusted gross margin was lower than normal at 41.3% for Q3 2025. Q-on-Q decline was due to a higher contribution from wire bonders, lower TCB revenue and a relatively lower manufacturing utilization in Q3 2025. Year-on-year decline was due to high base effect from TCB manufacturing ramp in Q3 2024 and a higher contribution from wire bonders this quarter. Encouragingly, year-to-date SEMI adjusted gross margin has stayed in the mid-40s and AP margins have remained stable.

SEMI adjusted segment profit was HKD 82.6 million in Q3 2025, down 52.8% quarter-on-quarter and 41.5% year-on-year, mainly due to lower gross margin and higher operating expenses, as mentioned in the previous slide.

Next, the SMT Solutions segment of our business. SMT delivered strong revenue of USD 227.5 million, up 28% quarter-on-quarter and 14.6% year-on-year. This was due to a robust performance in Asian markets driven by AI servers, EVs in China and the delivery of a smartphone bulk order booked in the previous quarter. However, contributions from automotive outside China and industrial remained soft.

SMT registered Q3 2025 bookings of USD 254.7 million, down 5% quarter-on-quarter, but up 51.8% year-on-year. Marginally lower quarter-on-quarter bookings were due to a high base effect from the Q2 smartphone bulk order, while the year-on-year increase was driven by strong momentum across both AP and the China mainstream markets.

AP bookings were supported by demand from IDMs and OSATs for telecom base stations and AI servers. China's mainstream business recorded strong year-on-year growth due to demand from EVs. SMT delivered a gross margin of 33.9% this quarter, up 136 basis points quarter-on-quarter and 163 basis points year-on-year. And segment profit was HKD 163.0 million, up 205% quarter-on-quarter and 65.6% year-on-year. Both were driven by higher volume effects.

With that, let me now pass the time back to Robin for Q4 revenue guidance.

Thank you, Katie. Now to Q4 revenue guidance. The group expects Q4 2025 revenue to be between USD 470 million and USD 530 million. This is up by 6.8% quarter-on-quarter and 14.3% year-on-year at the midpoint, which is above market consensus. This growth will be supported by momentum in both SEMI and SMT.

Looking ahead, the group's TCB TAM projection has a potential to go beyond USD 1 billion in 2027, supported by recent news about investments in the AI ecosystem. AI data centers will continue to drive demand for AP, particularly TCB for HBM4 and advanced logic where the group has technology leadership.

The group's mainstream business will be supported by global investment in AI infrastructure and stable demand from China, while visibility for automotive and industrial end markets recovery remains low. While the group has not experienced any material impact from tariff policies, it acknowledges that uncertainties remain. The group's global presence will provide flexibility to navigate any potential impact, and we will continue to monitor the situation closely and adapt as needed.

This concludes our third quarter 2025 presentation. Thank you, and we're now ready for Q&A. Let me pass the time back to Ben to facilitate.

Thank you, Rob. [Operator Instructions] With that, may I request Gokul to unmute.

First question I had is on the HBM4 commentary from you, Robin. And you mentioned that you are leading this transition to HBM4. Could you explain a little bit more what exactly that is indicating? Do you think that you would have higher market share in HPM4-based TCB compared to the incumbent Korean vendor?

And also your updated view on when does the fluxless TCB insertion happen for HBM? Is it happening for HBM4 or we are waiting for HBM4E for this migration to happen?

Gokul, have you finished?

Yes, that's my first question.

I think the question is on the HBM4, right?

Leading transition to HBM4.

That's right. Yes, I think as mentioned in our MD&A, we believe we have established ourselves as a primary supplier for the HBM4 market. I mean we have a conviction because we have won -- we are probably the first to have won the HBM4 orders for not just 1, but 2 major HBM players.

Now I think the second question is on fluxless. We believe that there is point as the industry continue to stack higher and higher and move from HBM4 to [ 4E to 5 ], in our opinion, it's quite inevitable that they have to move to a fluxless solution because the number of IOs will continue to increase, the pitch will probably narrow down, the chip gap will get smaller. So all this means that fluxless will be a better solution compared to a flux-based TCB solution.

So just to clarify, Robin when you talk about 2 HBM vendors, does it include the biggest market share player? Because I thought they are still using the incumbent vendor, right?

Yes, of course, of course, as we said, we have won orders from 2 of the 3. So definitely, yes, we are talking to the leading one.

Got it. Understood. Also maybe next question is, I think you observed some pause in AP and TCB in Q3. What is the reason for that? And given your guidance for 7% Q-on-Q growth for Q4, could you talk a little bit about how SEMI's overall and TCB within that will be growing? That would be outgrowing that 7% or it will be growing slower than the 7%?

I think in terms of when we talk about pause, actually, it's really largely driven by the timing of key customers' technology road map, right? So we are confident that when they launch the new architecture, we will get the orders. So it's a matter of timing in our opinion. So TCB demand, whether in terms of booking or billing will actually align with this timing as far as concerned. So it tends to be a bit lumpy, yes.

Is that more about logic? Or is it more about HBM? And also any indications on like segment-wise too, how are we thinking?

Gokul, I would say both because the technology road map will drive both HBM as well as on the logic side as well. Yes.

Okay. And Q4, any thoughts?

I think in terms of, if you are alluding to booking, maybe it's time for me to give you some color on booking for Q4, right? I'm sure this question will pop up during the conference call as well. Now the way we look at Q4 booking color on the group-wise in Q4, group-wise booking in Q4, we expect our bookings to be kind of flattish compared to Q3 reported number of -- Q3 reported number was USD 462 million. So going forward in Q4, we expect that to be kind of flattish on a group basis.

However, we do expect that this Q4 booking for the group will be the seventh consecutive quarter of year-on-year booking growth since Q2 2024. So it's encouraging to note that we have been growing our bookings for 7 consecutive quarters. And I think Gokul, encouragingly we see SEMI bookings are expected to increase by mid-teens Q-on-Q mainly due to TCB.

So I'd say Q-on-Q mid-teens and still comparing against the reported number, right? So that's for the SEMI bookings, expected to increase mid-teens Q-on-Q mainly due to TCB. So we expect TCB booking to sort of increase on Q-on-Q basis compared to Q3. And for SMT, we expect SMT to decline Q-on-Q due to the already high base effect in the prior quarter.

Now baked into the Q4 booking for SEMI, I think the [ QR ] for chip-on-substrate application. And as the market moves towards larger compound die because of higher computing power -- compute requirement, we are confident of achieving a sizable TCB order for OS application in Q4 from the leading foundry OSAT partners. And these orders will be likely built in early part of 2026, which will be definitely gross margin accretive, right?

So I think to sum it all in terms of Q4 and certainly beyond Q4 in terms of booking color, we remain confident that the strong AI tailwinds, including the recent news regarding investment in the entire ecosystem for AI will continue to drive demand for AP, in particular, our TCB technology leadership will position us strongly going into 2026 and beyond. So this is a bit of Q4 color and slightly into Q1 as well.

And next, I would like to request Donnie to unmute.

My first question is a housekeeping question. So considering we have disposed the AEC operation in China, wondering if Katie can give us some color on how should we estimate the OpEx or OpEx ratio in the coming quarters as we have seen the OpEx ratio has been pretty high for the past few quarters. So wondering if it will be coming down after the disposal of the AEC operation and also some cost control management.

And my second question is regarding to the TCB. So my understanding is that despite we have some progress in fluxless TCB, but the real volume shipment remains small into maybe fourth quarter this year. So I was just wondering if you can give us a time line where -- when exactly the fluxless TCB for memories and for leading foundries can ramp up more significantly in the future.

And also some comment on the progress in China will be also appreciated. As you know that China has been aggressively increasing their AI chip production capability, including HBMs as well.

I think I will take the first portion, Donnie. So you asked a question about AEC liquidation. I just want to make a correction. For AEC liquidation, as we announced, the savings was going to be RMB 150 million each year. Majority of that saving actually will be benefiting COGS, not OpEx. There will just be a little bit factories and G&A that will be part of OpEx. So that -- so on AEC, let me just spend a quick minute. The liquidation took -- sorry, the announcement took place in August, and the project has been progressing pretty well. And we do expect that the savings will benefit us going forward.

And on the OpEx ratio and specifically on OpEx, there's actually no change. At the beginning of this year, we announced that we will be investing incrementally HKD 350 million in R&D, especially AP and the infrastructure of the company. So every quarter, we've been actually -- we are on the path of the investment. And because of that incremental investment, we've mentioned in prior quarters that this year's OpEx will be similar to prior year with some marginal increase. And that narrative has not changed and will not change for the year.

Okay. I will take on the second question, Donnie. In terms of TCB fluxless application. As mentioned in our MD&A, we have made very good progress in terms of fluxless [indiscernible] TCB for logic side, Chip-on-Wafer. I think plasma technology has been endorsed by the leading foundry.

And also just to recap, Donnie we have been saying already in the past, but it's good for recap that Chip-on-Wafer demand this year, even we have won the technology battle, the Chip-on-Wafer demand will not be significant this year. We are looking into 2026 for inflection point in terms of Chip-on-Wafer application for logic TCB fluxless. Now I think that's your question, if I'm not wrong.

Yes, next question on the time line for the memory and leading foundry shipment.

Yes. I think in terms of fluxless, I answered the first question to Gokul already. So I think it all depends on when they will adopt the fluxless TCB for memory. As I said in our opinion, as the industry continues to step higher, the chip get smaller, more IOs in our opinion, at some point, quite inevitable that they have to move towards a fluxless TCB solution even for HBM.

And any color on China's adoption of TCB or opportunities there?

Yes. Donnie, I think we have been saying we are -- we supply to the global customer base. I think in terms of volume, obviously, the rest of the world volume in TCB is still higher than those of China. And for sure, we -- China ambition to really step up in terms of advanced packaging.

And next, I will request Kevin to unmute.

My first question is on the TCB outlook. As mentioned on the logic side, we are already passing the qualification, right? So I was wondering how should we think about the potential business opportunity on the chip-to-wafer part as compared to chip-to-substrate, as mentioned that most of the contribution will be coming from next year. And when is it likely the timing of this contribution will start?

And also on the memory side, I think we just mentioned that HBM4 we are screening order from multiple customers, right? So just wondering for the customer, are these for sample tool or for production already?

I can answer your first question first, Kevin, in terms of chip-to-wafer. Quite similar answers to Donnie. Chip-to-wafer in terms of volume, we expect it to be still smaller compared to substrate because substrate, I think the whole industry has moved -- almost the whole industry has moved to TCB solution. Whereas for chip-to-wafer at the moment, it's only the leading foundry leading the pudding in terms of using a TCB for particular end customer.

So if more end customers adopt TCB, then you will see Chip-on-Wafer TCB solution fluxes will increase. Otherwise, it's just one customer. I think the volume will still be smaller than the substrate volume.

Now in terms of HBM4, I would say they are already into some kind of a small volume production already using our tools for HBM4 production for the 2 customers that we talked about.

My next question is on the hybrid bonder side. So we are -- I was wondering how competitive are we in our Gen 2 hybrid bonder, which [indiscernible] we are already shipping? And what kind of chip order process are these for? Or this is going to be for mainly on the logic side or for the memory side?

Yes, Kevin, we have -- I would say we are shipping HBM -- HB, hybrid bonding solution for both logic and memory. As we speak, we are actively collaborating with other key logic and memory players and we're making good progress and all these projects are at different stages of evaluation. So we are hopeful that at some point when the hybrid bonding market takes off, we are there to compete with incumbent.

Okay. So we have already -- are we securing order from these customers already? Or this is just right now still in the evaluation process?

Yes, still in evaluation for some of these very key logic and memory players. We are engaging them very actively as we speak.

And next, I would like to request Sunny to unmute.

Could you hear me okay? So my first question is on a high level, directionally, how should we think about the recovery of mainstream SEMI solution from here? I wonder in the last few months, now given more manageable impact from tariffs, do you think the overall client sentiment is improving or not much change for 2026?

Thanks, Sunny. I think in terms of mainstream, I would say quite encouraging because mainstream are now also -- I mean AI also contribute to the mainstream demand. I think as you're probably aware, China is a significant portion. So we see China volume has been picking up for the last few quarters. So that's giving -- that's supporting the mainstream quite a fair bit for both SEMI as well as SMT.

Now in terms of tariff, I think the initial part of the year and initial period of the year, I think the tariff situation definitely has some impact on the sentiment of our customers. Now I think with the tariff situation a little bit more stable, I think customers are now a little bit more confident, I would say, in terms of placing orders. That's why we are also seeing -- we have good orders coming from mainstream wire bond, die bond and SMT are also seeing a mainstream application for putting chips on larger PCB boards for base stations and all that. So all these are also partly driven by the AI adoption.

So in general, we see mainstream certainly coming up on the bottom. But going forward, we see mainstream stable, especially the demand coming from China provides that kind of stability for mainstream.

Got it. And then I have questions on TCB. Maybe if you could remind us the lead time for you to make TCB tools nowadays. In terms of orders, should we expect the inflection point to potentially come maybe in first half or second half of 2026 for logic and for HBM?

I think for logic, I think we mean that sizable orders for the chip-on-substrate for larger compound die will most likely realize the revenue in the early part of 2026. For HBM, it all depends again on the timing of our key customers' technology road map. So if they accelerate, we will see revenue earlier for HBM. If there's a further delay, then our timing will also align accordingly.

Now in terms of TCB lead time, actually, internally, we are efficient. We don't take a long time to assemble a TCB machine. It all boils down to material supply, right? So if we -- if customers give us more visibility, we can order materials earlier, then the lead time will be shorter. So I think that's the dynamic of the TCB lead time at this point in time.

Sorry, maybe a quick follow-up. So for logic -- so on Chip-on-Wafer, any view on when the leading foundry may start to migrate to TCB? Maybe will that be in second half of next year or early 2027? And therefore, assuming if your lead time is about like 2 quarters, should we see orders starting to come through maybe from first half of next year?

We are hoping orders will come sooner. But again, as I say, it depends on the timing of the road map. We are confident that chip-to-wafer, we will have delivery or shipment in 2026. I don't think it will delay to 2027.

And next, I would like to request Daisy to unmute.

My first question is for Katie regarding the SEMI Solutions gross margin. Katie, you previously mentioned that the closure of AEC will have a positive impact of the cost of goods sold going forward. Yes. So how we should think about the near-term and the long-term gross margin for the SEMI Solutions segment?

Daisy, assuming you're kind of talking about basically the gross margin going forward, right?

Yes.

Okay. So first on the AEC point, is correct. We would expect the savings to come in gradually in Q4 and then full-fledged in next year. Now in terms of the overall SEMI Q4 gross margin, we do not provide guidance, but just some kind of directional pointers. Robin guided Q4 revenue probably could tell that the TCB contribution -- revenue contribution will continue to be lower, but with some have high photonics but wire bond momentum will be sustained. So therefore, we expect a slight margin accretion for SEMI's margin in Q4.

And then when you look at the group level, then if SEMI and SMT mix stays similar and SMT experiences a stable margin, then we expect basically slight margin accretion for the group in Q4. Now of course, we always caveat right it's really depending on the mix going forward, especially in the midterm in kind of longer run, we -- the technology leadership in HBM and advanced logic with those leadership, we expect the TCB order in Q4 and beyond -- I'm talking about in the midterm now, would actually provide support to SEMI's gross margin. And with this liquidation that you mentioned earlier, we do expect that the SEMI gross margin will come back to the kind of the mid-40s level.

It's clear. And second question is for Robin on the hybrid bond. So you are at an evaluation stage for the leading foundry and HBM customers. So for the HBM use hybrid bond, do you see that it will happen in 16-high or 20-high.

Since we are a dominant TCB player, we hope that they can continue to use TCB even up to 20-high. But nevertheless, we are prepared that if they have to switch to hybrid bonding, we will be there also to provide competitive solution for hybrid bonding for HBM 20-high.

Yes. And also a quick follow-up for your leading foundry customer, your European peer has been a dominant supplier for hybrid bond at that leading foundry customer. So how you see your hybrid bond opportunity at this leading foundry customer?

Yes. We will be relentlessly knocking on their doors for sure. But I think having said that we also have been saying that because we are not the leading player in hybrid bonding, I think the advantage is that we know the pain point, existing pain points, right? So with that coming in from behind, we are relentlessly and diligently working with all the leading logic and memory players, asking them what are the current pain points so that we can incorporate features, engineering innovations to mitigate or totally eliminate those pain points using our tools. This is what we have been doing.

So I think we are confident that our Gen 2 and in future Gen 3 should be able to address all the pain point and give us an entry point in all this leading key logic and memory players.

And sorry, final follow-up. So in the Gokul's question, you said that you are the primary supplier of the HBM4 market and the first company won the HBM order at 2 key customers. So is it the fluxless TCB or the flux TCB?

It's still the flux TCB at this point in time.

[ Flux 1 ]?

Yes. The Flux 1. Yes.

Next, I will request Leping to unmute.

I have another question about the TCB. So what are the current customer concentration level of your TCB equipment now? And what may look like in the future? So is it mainly still concentrated on the top 3 memory maker and the leading foundry? Or you also see some broadening of your customer to other OSAT or other foundries in the market? This is my first question.

I think we have definitely we have broadened our TCB fan-out to not just leading foundry, the OSATs, HBM and also globally as well. So we're pretty engaged with all top AI customers needing requiring or requiring TCB solution. I hope I answered your question.

Okay. The second question is about this -- you have the deposition equipment cancellation. So is it due to some the road map change of the -- in the advanced packaging? Also, I remember, is it due to the -- you have a company subsidiary called NEXX, it is from that subsidiary.

It is from that. It is from NEXX. I think it's a case of digestion of capacity, right? So there was a bit of a sizable capacity maybe about 2 years ago, right? So the customer take time to digest. So -- and these particular customers decided to give it up and pay us a cancellation fee.

Next, I will request Alex to unmute.

First question is about your margin on SMT solution. It seems like your quarterly revenue level already increased to the level similar to 4Q '23 or early '24. So I see the margin still like low 30s to -- is this the normalized margin going forward? Or you expect margin can return to high 30s level sometime in the future?

Yes, Alex, this is Katie. Thanks for the question. So for -- you're kind of comparing to a few years ago where actually the SMT's end market composition were quite different. The -- few years ago, actually, automotive and industrial were running really, really strong and their contributions to SMT's revenue were much larger. And this is where we actually could command relatively higher margin.

So currently, as we mentioned, the automotive and industrial end markets are relatively muted. And that's why the margins are sitting in the, call it, low 30s. Unless the end market composition changes, this kind of level will be sustained in terms of margin percentage.

Another follow-up question on TCB. You mentioned the TAM would reach like USD 1 billion in 2027. Do you have probably a rough split between the logic versus memory and also split between C2W applications?

Yes. I think it's dynamic. I would say, Alex, it's very dynamic. Again, it all depends on customer road map and all that. But generally speaking, if you take really looking further into the future, it's just intuitive that the HBM TCB demand or size or TAM will be larger than logic because of the number of stacks and also as the industry migrate from one architecture to the next generation, they require more HBM stacks per chip, right? So naturally, I think HBM demand over time, not in a particular year, not in a particular quarter, but over time, HBM demand for TCB will be larger than logic.

Got it. So what is the company's target market share for each application?

We don't go down to that kind of granular level HBM market share or logic. But overall, I think last year, we put out the TAM for TCB, our aspiration is to hit 35% to 40% market share in the entire TCB TAM.

And next, I would like to request Arthur to unmute.

Can you hear me? So the first one, Robin, if you can -- can you share with us a high-level ballpark figure on the revenue contribution from AI?

This is a difficult question. I think for -- we don't share -- sorry, first, we don't share, but also this is a difficult question because we talk about AI benefiting both AP and mainstream. Well, we have better visibility on how AI benefit AP. But in terms of mainstream, it's a little bit tricky because wire bond, die bond, they are quite fungible. Today, customers may say, okay, I use it for AI-related packaging, tomorrow, they use it for others. So it's a bit difficult to really unpack, sorry.

No problem. Because you just mentioned that you saw some power application, they start to come back and the drivers from the AI. So that's why I want to get this high-level ballpark figure. Maybe we can discuss it next quarter when we have a visibility.

The question number 2 is on the cancellation from the high-density substrate. And I think Leping already touched base a little bit. So my question is, is the key component of the equipment fungible? Can you give it to the other substrate customer?

The short answer is yes, there's no inventory related issue relating to this cancellation.

Because if we look into the AI business of the rack and also the key component, actually, we heard more and more PCB, HDI substrate shortage at this moment. So I'm kind of wonder, so was the client is based in Japan or in Taiwan or China?

None of this actually, none of this. I mean this is a NEXX business, we are saying that they supply to a few key players, high-density substrate players. It just happened that, as I said, I repeat again, it just happened that there was a big capacity ramp-up in the last 2 years. And this particular customer just say, okay, I'd rather not keep you holding on all these orders, I decided to cancel it. So I think in short, this is that kind of circumstances.

So in the future, when we look back, this could be an isolated event. So do we think this demand for the other customer will return?

No. This in a way I don't -- if you are thinking is this AI related, I wouldn't say this is AI related. Yes, this is -- they are serving a particular IDM which use all this equipment for RDL and all that. So it's a particular application. I would say it's not related to AI. So don't link this cancellation with AI that we have been talking about. De-link these 2 pieces, Arthur. This cancellation has nothing has nothing to do with AI.

I think we have time for one last question. I think we have Gokul here.

So my question is more on the margins and operating leverage. I think we are having pretty good momentum both now in mainstream and in TCB. Margin still seems to be a little bit sluggish. How do you think, let's say, next 2, 3 quarters, TCB revenues will come through given all these orders, bookings realize into revenues. What does it do to gross margins? Like is TCB still accretive to group gross margins right now? Or is it kind of similar to group gross margins?

Second part of the question, again, to Katie is on operating leverage because now that we are back to some degree of revenue growth, we're still not yet seeing meaningful operating leverage come through. I'm asking because Street expectations are for very big operating leverage to kick in for next year. I think revenue growth of 10%, 15% contributing to doubling of your operating profit is what a lot of Bloomberg estimates are looking at. So just wanted to understand what is the extent of operating leverage that we can expect?

I think we have seen operating margin go back to high teens to 20% at really, really peak kind of levels back in 2021. But in the recent past, we've not really seen operating margin really get beyond the mid-single-digit levels. So just wanted to understand what is the extent of operating leverage we can expect as we start some of these ramp-ups for TCB and other products?

I appreciate the question. First thing, TCB, I just want to make it very clear that TCB margin has been stable and is accretive to SEMI business. Now overall, when you say operating leverage, volume has come back up, but not quite at the super cycle level. And within the volume, we always say there are a few mixes that actually impact margin. One is the segment mix.

So far, as you can tell, like in Q3, for example, the SMT contribution to the group is at about 50%, right? SMT naturally has lower gross margin. Therefore, the segment mix could be different based on the contribution from the 2 businesses.

The other thing is on product mix. Within SEMI, for example, it really depends on the product mix between TCB and wire bond in Q3 and as we guided for Q4, if you look at that product mix, when we have less TCB revenue, but more wire bond revenue coming from mainstream applications, the margin -- the gross margin side would not -- would be under certain pressure.

But having said that, in the long run, as a few of you asked earlier, we do expect that our SEMI business will continue to enjoy the accretive margin contribution from applications like TCB. And with the AEC liquidation we mentioned, we should have savings from operation efficiency, et cetera. So that I think our conviction for SEMI gross margin to stay in the mid-40s and then gradually going up has not changed.

And then so at the group level, we've been talking about the 40%, right? I think, again, I'm talking about in the long run, not a specific given quarter, I think we are comfortable that the group's gross margin will be at that level and gradually improve as we go.

Got it. So just on the OpEx side, same, because that's something that you can control revenue harder to control, especially on mainstream. Are we going to stay around this roughly HKD 5 billion kind of level going into next year? Or we still see that OpEx will keep growing given we are investing in some of these newer technologies?

Yes. So Gokul, I actually cannot answer your question very well right now. Maybe give us a quarter because the organization actually is going through the budget process. But directionally, as we have talked about before, the OpEx has been running at HKD 4.7 billion in the last few years. And this year, with the R&D and infrastructure investment, we have communicated that will be marginally higher.

Though the investment is at HKD 350 million, we are doing certain restructuring projects and the cost saving projects that you probably have seen in the last few years on trying to bring it down. So this year, I think you guys can do the math, right, it's about HKD 2.8 billion. So that's kind of where we are.

I think going forward, Gokul, we're not going to change our commitment in R&D investment as you guys were talking about TCB, hybrid bond, all that, that side of the conviction has not changed. We'll continue to do the right investment.

On the other front, for the overall efficiency and productivity of OpEx, we'll continue to look into any opportunities we can find and trying to contain that. So again, I cannot give you a specific number. We'll probably share with you more. But I think our strategy -- our thinking on OpEx has not changed.

That will be all for the last questions. And I will now pass the time back to Robin for his closing remarks. Thank you.

Thank you, Benjamin. Just a couple of pointers before we officially close the call. The group maintained strong business momentum this quarter. Our AP and mainstream business will continue to benefit from sustained AI adoption. TCB solution, we secured repeat orders in both memory and logic, reflecting ongoing technology leadership, particularly in HBM4 and advanced logic.

What Katie said, we are in the midst of really finalizing our budget for 2026. But certainly, I can -- at this juncture, we can give you some direction or some color on how we look at 2026. We expect a growth year in 2026 largely driven by AP because of AI and underpinned by the sustained momentum of our mainstream business. And finally, we remain confident in the total addressable market for TCB, which we believe could go beyond USD 1 billion in 2027.

So thank you. With that, we will close the call and see you next quarter.