The conference is about to begin. The senior executives present today include President Anson Chiu, CFO K.T. Lim. Today's agenda consists of three parts. First, CFO K.T. will explain the financial and business performance of Q2 and H1 2026. Next, President Anson will explain the company's operational outlook and growth strategy. During the Q&A session, Anson and K.T. will provide further explanations. First, CFO K.T. will explain the financial results of Q2 and H1 2026.
Good afternoon, everyone. First, I will present to you our results for Q2 2026. Lite-On Technology's Q2 revenue was TWD 52.7 billion, GP TWD 14.3 billion, with a rate of 27.2%. OP TWD 8.2 billion, with a rate of 15.6%. Net profit after tax TWD 7.1 billion, EPS TWD 3.14. Overall, Q2 revenue went up by 30% year-over-year, in which Cloud and AIoT revenue grew by more than 70% YoY.
Revenue went up by 21% quarter-over-quarter, with all three major segments showing double-digit quarter-over-quarter growth. GP went up by 60% YoY with a rate of 27.2%, up 5.1 percentage points YoY. OP went up by 120% YoY with a rate of 15.6%, up 6.3 percentage points YoY. It mainly benefits from strong demand for next generation AIDC. In addition, shipments of high-end models were deferred from Q1 to Q2, thus increasing the share of high-value businesses. Meanwhile, global capacity expansion brings economies of scale, while smart manufacturing optimizes operational efficiency, driving overall profit growth. Net profit after tax TWD 7.1 billion, EPS TWD 3.14, up 1.26 times YoY, a record high. Now, results for the first half of 2026. H1 revenue was TWD 96.1 billion, up 25% YoY. GP TWD 23.8 billion with a rate of 24.7%. OP TWD 12.3 billion with a rate of 12.8%.
Net profit after tax TWD 10.9 billion, EPS TWD 4.8, up 66% YoY. Compared with the same period last year, Cloud and AIoT continued its strong growth. In H1, revenue grew by 60% YoY. GP went up by 38% year-over-year, up 2.4 percentage points YoY. OP went up by 66% YoY, up 3.2 percentage points year-over-year, showing the expansion of high-value businesses and the improvement in both operational efficiency and profitability. Net profit after tax in H1 was TWD 10.9 billion, EPS TWD 4.8, up 66% YoY, a record high. Now very quickly, I'm going to talk with you about revenue and profit of the three major segments. Q2 revenue from the three major segments went up by 21% quarter-over-quarter, OP TWD 8.2 billion, doubling from Q1, in which in Cloud and AIoT revenue was TWD 28.9 billion, up 26% quarter-over-quarter, accounting for 55% of total revenue.
It benefits from the growing shipments of next generation high-end cloud power products. OP TWD 6.9 billion, doubling from Q1. In optoelectronics and also in ICT, revenue has also grown quarter-over-quarter and profit has also grown quarter-over-quarter. Overall, we continue to strengthen high growth, high value core businesses, increasing both the overall operating scale and profitability. Now, let's look at our balance sheet. In Q2, accounts receivable went up by TWD 7.5 billion in current assets. Inventory went up by TWD 11.2 billion QoQ. To manage shortages and price increases of critical materials, to meet safety stock requirements, and to manage global operations, we initiated strategic procurement and material preparation to strengthen supply resilience and delivery capabilities. In current liabilities, short-term bank loans went up by TWD 17.9 billion QoQ.
Accounts payable went up by TWD 8.5 billion QoQ. Q2 current assets went up by TWD 24.7 billion QoQ. Current liabilities went up by TWD 20.6 billion. Quick ratio and current ratio 0.94 and 1.39 times respectively. Net cash position TWD 40.8 billion, down TWD 13.8 billion QoQ. It was mainly used to pay cash dividends of TWD 6.8 billion and CapEx of TWD 4 billion in Q2. Starting in 2021, the enhancement of high-value businesses has driven high-quality revenue and profit growth. Since the rise of AIDCs in 2022, high-value cloud computing business has grown. The revenue share of cloud and AIoT went up to 54% in H1. It was only 32% back in H1 2022. Revenue scale has nearly doubled in size. H1 revenue went up by 25% YoY, and OP went up by 66% YoY.
High-value businesses have continued to drive excellent annual growth in revenue and profit. Next, you can look at our ROIC enhancement. Behind long-term forward-looking investments, our management focus is on AI infrastructure to generate stable ROIC. ROIC performance continues to show the success of our strategic transformation and capital allocation. Between 2020 and 2023, by focusing on high-value businesses and optimizing our portfolio, ROIC went up from 28%-55%. Our ROIC in the first half of the year reached 22%, and in the second half of this year, it's expected to be better than the historical average. It shows that while actively managing capital, improving efficiency, and expanding AI infrastructure capacity, the growth of high-value businesses not only drives revenue and profit but also creates long-term shareholder value.
As for our strategic deployment, in addition to R&D investment, in response to the trend of globalized manufacturing, we have successfully completed our manufacturing deployment in Taiwan, Vietnam, Thailand, and North America. In which Kaohsiung Plant phase II in Taiwan and Quang Ninh Plant in Vietnam are expected to officially start mass production in Q3 and Q4 this year respectively. With rapid business growth, CapEx for 2026 has been revised upward to TWD 18 billion from TWD 13 billion announced last quarter. These capacity deployments not only demonstrate our confidence in future growth but will also significantly enhance Lite-On's ability to secure orders from new customers and for new products. This is the overview for Q2 and H1.
Revenue TWD 52.7 billion, up 30% YoY and up 21% QoQ, in which cloud and AIoT grew by more than 70% YoY. GP went up by 60% YoY with a rate of 27.2%, up 5.1 percentage points YoY. OP went up by 120% YoY with a rate of 15.6%, up 6.3 percentage points YoY. These benefit from the growing demand for next-generation AI infrastructure. In addition, shipments of high-end models were deferred from Q1 to Q2. The growing share of high-value businesses, economies of scale from global capacity expansion, and the optimized efficiency of smart manufacturing have improved overall profitability. Net profit after tax TWD 7.12 billion, EPS TWD 3.14, up 126% YoY, which is a new record for Lite-On. As for H1, revenue TWD 96.1 billion, up 25% YoY. GP went up by 38% YoY with a rate of 24.7%, up 2.4 percentage points YoY.
OP went up by 66% YoY with a rate of 12.8%, up 3.2 percentage points YoY. Net profit after tax TWD 10.9 billion, EPS TWD 4.8, up 66% YoY, which is also a record high. Core businesses. Thanks to smooth shipments of high-end AI server power, cloud computing products, and power management systems, cloud and AIoT revenue doubled YoY. Next, BOD resolution from this morning, BOD approved a cash dividend payout of TWD 2.5 per share for Q2 2026. This concludes my presentation.
Thank you, K.T. Next, President Anson will explain the company's operational outlook and growth strategy.
All right. Thank you, Julia. Thank you to our investment partners and media friends. Good afternoon. Now that we've heard K.T.'s presentation, I'm sure you would all agree that Lite-On today is very different from the Lite-On of the past. This change is not just seen in those financial numbers that we talked about, but also in our role. Lite-On is no longer a company that follows the ODM or OEM model. We have transformed into a solution provider, providing high-value solutions to the market and our customers. This transformation comes from one core idea, which is that only through technical innovation and playing a leadership role in the industry can we continue to guide the upgrade and development of our industry. Lite-On's power supply products in the AI industry is the best example.
It took a very long time, a lot of cooking, from investment to actual results. As you can see today, all of these investments were very worthwhile. We will continue to follow this central idea, continue to invest in our products and technical innovations. With future trends in new energy structures, we will continue to strengthen our R&D deployment in North America and develop new products and new technologies where our customers are. This helps us provide maximum value and to maintain our leadership role in the industry. I talked about new products, and I'm sure you've heard a lot about them. I think, as you know, we have been very focused on the AI sector, particularly when it comes to power management.
Our new generation 110 kW Power Shelf PBU and 800 volt HVDC Power Rack, as well as our future high-end D2D Bricks, would all enter qualification and production in the second half of this year. In addition, when it comes to liquid cooling products, such as the 120 kW CDU and the Sidecar, all of these have entered mass production and shipment in the last quarter. We anticipate stepping up the volume in the next half of the year. When it comes to physical products, our rack level products have received orders from cloud service provider customers in North America. Through this varied and diverse product portfolio, we can serve our customers' needs and drive further growth momentum for the company. We anticipate that this year, our share of AI products in our revenue will exceed 30%.
At the same time, as we plan for the next generation AI data center market and the optical communication markets, our board of directors approved a strategic investment in DenseLight, a optical communication component company from Singapore. This will allow us to procure 21% of the company's shares and one seat on the board. DenseLight has R&D and production in indium phosphide products, and their core products include CWDM laser gain chips and SOAs. These are all key components in high-speed optical communications. Through this acquisition, we will use our existing advantages in advanced packaging, system integration, and global manufacturing to broaden our strategic planning in the optical engines, optical modules, and rack scale optical connection markets. This will help us seize the long-term growth opportunities in AI data centers and high-speed optical communications. Over the years, we have talked about our long-term strategy in AI infrastructure.
Starting from this year, we're not just talking about the strategy anymore, we're seeing results from implementing these plans. The second quarter was an important milestone for Lite-On. The revenue and profit that you saw comes not from short-term market opportunities, but from long-term investment into R&D, global production, and talent development. This has now borne actual fruit, and it has shown that Our forward-looking investments have been transformed into tangible growth momentum. In the future, we will continue to deepen our strategic planning in power supply, storage and transformation, cooling, and rack solutions. AI will continue to be our core growth engine as we work toward becoming an important collaboration partner for AI infrastructure around the world. That concludes my presentation. Thank you.
Thank you, Anson. Now is the Q&A session. We open the floor for questions, and Anson and KT will provide further explanations.
Hello, Anson, KT, and Julia. I'm Joyce from BofA. My first question is this. In Q2, in the gross margin. I remember that in Q1, there was a one percentage point inventory write-up. In Q2, was there a reversal of that? This is my first question. Do I continue with my questions? Okay. My second question is this. This is about CapEx. This time it goes from TWD 13 billion-TWD 18 billion. Does it include the TWD 900 million investment in Texas? I remember that the TWD 13 billion didn't include that yet. What about the TWD 18 billion? Is it included? Do you also have some other expansions? Does the expansion include BBU expansion? Because the demand is also quite strong. Here's my last question.
In the second half of this year, we see that some critical materials and parts and components are in shortage. Do you have some strategies to address these challenges? When the cost goes up, will you be able to respond to customers? As for CSPs, you have two big CSPs as your targets, right? Has there been any progress that you can share with us? Thank you.
Let me answer your first question. Well, in Q2, when customers were pulling goods, there was some reversal, it didn't affect Q2 that much. As for your second question related to CapEx. Yes, when it went up to TWD 18 billion, it included the planned expansion in the U.S.. Yes. Let me answer your last question. As we told you previously, the North American market is an important one for Lite-On. Our products don't just service one single customer. All the CSPs in North America are our targets, and we have made efforts in this direction for many years. This year, we can happily tell you that we are already delivering products to those two CSPs. The volume is not big yet. In one year, we should be able to establish our credit in their records, and we should continue to increase our market share.
Your third question is about our products. Will we continue to invest in AI-related products with our customers? What I want to tell you is this. When it comes to R&D, we will continue to increase our R&D percentage, which is currently 5% to 5.5% in the company. In the future, we will be able to continue to increase our R&D so that our product mix in the company can continue to satisfy customer needs in the future. Well, when it comes to materials, as K.T. told you earlier, in Q2, our inventory went up. I think, well, to answer your question, there are risks of price increase and shortage of materials. In Q1 and in Q2, we had strategic material stocking, and our customers were informed, meaning that they were willing to provide concrete subsidy or guarantee for these material stocking.
In Q2, our inventory went up, these materials are enough for our critical parts and components in the second half of this year. When there's a price increase and shortage of materials from suppliers, I think we have good plans to manage the situation. Rest assured, in H2, these shouldn't be the obstacles for our growth. To add to what has been said w hen it comes to materials, every month we have business development projections. We told you about our strategic procurement, which is done six months in advance. When it comes to materials, there are three different management methods. For new products, we more actively procure in advance. For each product and its upstream and downstream, we have very stringent management.
Hello, Anson. I want to ask a bit more about the BOD investment resolution. You said that this company makes indium phosphide, right? Well, I'm not very familiar with this industry, so could you tell us more about DenseLight and their architectural design and wafer process? What's their market position? If it's a startup, is it losing money or is it breaking even? What kind of a company is that? I would like to invite the manager in charge to provide more detailed explanations.
Hello. This company has made indium phosphide for 20 years, and also architectural and wafer process. It's an IDM doing high-end laser. Why do we do this? Well, because in the long-term growth of optic communications, technology iterations are important, and solution integration is also key. Lite-On has our long-term optic packaging and testing capabilities and assembly capabilities for a long time. Also we have photocoupling in optic communications. Indium phosphide now has some supply gap where we need some strategic cooperation and operation. This company is a robust company. We hope that Well, this company has good technology. We want to develop with them for 1.6T and 3.2T light engine or CPO solutions, basically next generation optic communications. We look forward to this.
Were you talking about 2025? For 2025, I mean the company. What is the company's financial state right now? Can you discuss that?
Well, this company is currently a private company. It's a startup private company, and we will own 21% of the shares of this company.
Well, I remember that, Anson, you talked about some strategic planning surrounding CPOs. Is this investment part of that?
Yes, it's part of that. As I talked to you about when it comes to the optical communication market, we will be focusing on optical engines and optical modules. You can think of what this company produces as a chip, and Lite-On's strength is in packaging. It's also a vertical integration play. Through Lite-On's strengths in packaging, we can help this company achieve higher competitiveness in production and in shipments. In addition to their technology, we can provide synergy when it comes to production for this company.
Your investment into this company is to take their indium phosphide laser and make it into a component of the CPO. Is that the play?
Yes. We're going to make their chip part of our module, to make it the light source for our optical communication product.
Anson Chiu, K.T., Julia, good afternoon. My first question is about the gross profit margin, because it went up quite significantly this quarter. Can you talk a little bit about why? It seems that every business unit saw growth in the gross profit margin, with the highest coming in the cloud business. It seems that the growth of the cloud business has been in line with previous quarters. Why has the margin suddenly gone up?
Well, if you remember our investors' conference last quarter, we talked about gross profit, GP, and we mentioned an inventory write-off. That was why the GP was relatively low last quarter. In addition, one of our customers had a delay in their shipments, and because of this delay, the SKD expense did not come in until the second quarter.
If you balance these various factors in the first and second quarters, our overall gross profit margin for the first half of the year is about 24%-25%, and I think that is a realistic target for the next coming quarters as well. To add a little to that, the quarter-to-quarter volatility has various factors behind it, because it might result from shipping delays, and the shipping delays might result in changes in our product portfolio. I think it would be better not to focus on quarter-to-quarter volatility. The overall 24.7% profit margin for the first half of the year is what should be focused on. We talked about a target 25% profit margin in the past, and I think that is the number to focus on when evaluating Lite-On's long-term strategic planning.
If we look ahead at the second half, 24%-25% is a more reasonable target, or will it be 27% again?
We will always be working to improve our gross profit margin, but I think long term, with the uncertain factors ahead, including materials costs and product portfolios, there are some uncertainties. I believe that our profit margin will hold steady at 25% in the future. I believe that is a realistic goal for us.
For the second quarter, were price increases a factor? We know that there have been price increases for some consumer products. Has this benefited your profit margins, and will the price continue to go up in the second half?
Well, I think all of our cost increases so far have been passed on to our customers, but our customers are very willing to accept these increases to secure their materials. As I said, some of them even want us to strategically procure more materials because they might be worried about further price increases or even disruptions to the supply. The material costs are going up, but our prices are also going up, the margin will hold steady. It's holding steady at 25%? Not 27%, correct.
Next, I would like to talk about the two new CSP customers that Anson discussed. You said that now they are receiving shipments, but what products are they receiving shipments of? Are they power supply or rack products or something else?
Mostly power supplies and racks. For the power supply, a lot of it is BBUs, and for the racks, it's the Power Rack product. Julia, can you go back to the slide about products?
I remember that in the last conference, you talked about the 800V DC, and it was quite clear that it would ship in the third or fourth quarter. This time, you said that it would complete qualification. Does this reflect a change in the timeline or a change in the market?
From our perspective, this isn't a change in the timeline. Our current progress, we can't comment on our customers' progress, but on our part, the 800V product will have some sample runs starting in August. Part of that will go to our customers, but part of them will be here. We will conduct a 13-week test on them. That's the qualification process I'm talking about. After that, the Power Rack will be integrated with the IT system of our customers, and that's a second phase of the qualification. After that second phase, if that goes smoothly, we can start small-scale mass production in November. For larger-scale mass production, it would probably be the first quarter of next year.
Okay. When it comes to the 800V DC product, it's a specification that your GPU customer has been promoting. That GPU customer is not the ultimate buyer of this product. Once you enter large-scale mass production in the first quarter of next year, do you already have real buyers lined up to purchase it?
Well, the HVDC product is not just for our GPU customer, it's also for our ASIC customer. The timeline that I talked about is actually mostly for the ASIC customer. They have been making faster progress. As for our GPU customer, they're also working on it, but their timeline would be a little bit later, probably the second or even third quarter of next year. We believe it's most likely going to be the second quarter.
Okay, I see. My last question is about DenseLight. Can you talk a little bit about their laser products? What milliwatt level are they working in now? Are there any exclusivity agreements between these two parties, whether it comes to production or procurement? What is their production capacity currently? At what time do you foresee this investment to start contributing to your revenues? If it's successful, will this be reflected in your opto business, or are you just an OEM for DenseLight?
Will the products be marketed under your branding or DenseLight's?
Well, we're the module producer. This is not an OEM agreement. Their product will be part of our module, but we will be separate companies. We're only owning about 21% of them, so their financials will not be part of ours. I mean the module.
Will it be part of your module?
Yes. Their product will be part of our module. As for the milliwatt question, I believe it's about 200 mW. Is it 100mW or 200mW? Well, currently, their mass-produced product is 100 mW, but we know that in the future, for the CPO and optical engines, 200mW and 400 mW will be the mainstream in the future, and we are on our way to those levels. Currently, their mass production is at 100 mW. We are focusing on the back-end optical engines and CPOs.
What is their production capacity?
Well, their production capacity is ever-expanding. To answer two of your questions, the production capacity is ever-expanding. There will be more machinery in the third quarter of this year and the first quarter of next year. In the first half and the second half of next year, we will see significant production expansions for that company. You also asked, what did you ask about? In the production capacity expansions, is there exclusivity? Well, there is not, but we do have many collaborations in terms of strategic products and production, and we want to bring together our strengths in the future products such as the optical engines, and we want to optimize all of them.
Do you already have end customers?
Yes, we have a few for design and development.
When will Lite-On see revenue?
I think it depends on the whole supply chain. I think in 2027 or 2028 will there be clear numbers. Why does it take so long? I think 100 mW is mainstream, so there should be many customers. When it comes to Lite-On CPO architecture, I think only in late 2027 or early 2028 will there be numbers. For 100Mw , they already have some business numbers, but that's their business, not Lite-On's business.
You only do ELS, you don't do CW laser optic modules?
We are discussing that. Optic modules doesn't take that long. Yes, we are discussing it.
Thank you.
Hello. I have a few other questions that I want to follow up with the managers. You mentioned your AI percentage, which will surpass 30% this year. Can you also share with us the Q2 percentage? This is my first question.
For Q2, it already surpassed 25%.
Understood. You also mentioned the two new CSP customers, and there will be some products that will be shipped. Do you target the new generation from H2 this year, or do you ship from the existing generation?
Well, we are already shipping, including power supplies, and also racks. They are being shipped already, for future customers, we will start from new models. There are already some products, but not very comprehensive. For the new generation, we are already in their AVL. We are developing new generation things now.
For new customers' shipment, ship or market share, do you have a target?
In the long run, we hope that for key accounts, in terms of their future revenue, we want to achieve $1 billion as our target.
Got it. Thank you.
Hello, managers. Sorry. I meant U.S. dollar, not TWD. Hello, managers. I want to ask you about CapEx. Your revision up to TWD 18 billion, you have continued investment in the U.S., I want to know, will the number continue to go up next year? What will be the pattern like? TWD 18 billion is quite a big amount in relation to your profit. Will that affect your dividend payout and do you need to raise capital?
In terms of CapEx, I think currently there won't be a chance that the number will be higher than the number this year. This year we have two major expenses. On the one hand, when we are developing high-Power Racks and these new products, in terms of R&D equipment and production, we need new things. About one-third of our investment is used to buy these R&D equipment and production testing equipment. This is the first item. The second is production line expansion. As the CFO mentioned, in Q3, our Hai Phong in Vietnam and Quang Ninh in Vietnam, and also our Kaohsiung phase II will start their mass production. Our S&D and production line investment is our second-largest expense. Once they are ready, in the short term, our CapEx will only be fine-tuning to satisfy customer needs rather than such a big CapEx this year.
I have another question about Q3 and Q4. What's your outlook for revenue and business segments? Thank you.
I'm quite optimistic personally.
We have a few questions for you. First of all, I remember that last month, you bought Nanjing Nenglixin Technology, which does DC-DC converters. Could you share with us what's their current revenue, customers, product R&D, and future synergy with you, as well as future customer strategies? This is the first question. Second, capacity expansion. You mentioned that in the U.S. and other places. For this year and next year's PSU BBU capacity, what will be the year-over-year growth? What is your take on that?
The third question, do you have 2027 overall revenue outlook? I know that it's a bit early still. Last question, in Q2, GPM was pretty high. Is the new platform related to Power Racks? The GPM of Power Rack, is it much higher than other products? What's the reason behind Q2's high GPM?
As I said previously. To answer your last question, in Q1, we had the write-off, some customers, because of material shortage, had to defer their demand to Q2, the product mix was different. During our presentation, we said that in Q2, cloud high power grew by almost 70%. Due to write-off and product mix difference, our GPM in Q2 was higher. This is why we said that when we look at the first half of this year, things look more accurate.
As for DC-DC brick, HVDC application will become an important component, and the most important thing is that the voltage is going to increase, and we will only focus on high-end DC-DC brick in our deployment and strategy. Nanjing Nenglixin Company is a technology-centric company. It doesn't have a lot of revenue there, and their production is currently mainly in China, but their technology is recognized by CSPs. We can cooperate with this company technologically. When a company is small, their technology cannot be easily used by big companies. We can bring their technology and present it to big customers. We believe that for customers, this serves as a sufficient guarantee. What's another question?
Well, I think when it comes to production capacity, we cannot comment on specific products. Overall, comparing 2026 to 2025, our overall production capacity outside of China has actually grown by 162%. We forecast that in 2027, production capacity will continue to grow 63% year-over-year compared to 2026. By the end of this year, we anticipate that our non-China production capacity will reach a 60% share of the total, and this will grow to 70% by next year. That is a brief overview of our production capacity. Thank you.
Well, I would like to add a bit. Last year and this year, I think the biggest issue with our capacity was that our new production sites were not yet ready, so we are still capacity constrained, and that's why we haven't seen very significant growth in revenue.
As these investments bear fruit, whether in Vietnam, in Taiwan, or in North America in the future, all of this capacity will lead to a near doubling of our current production as long as we continue to invest in equipment because the building and the land is no longer an issue. These investments will be very important to the next five years when it comes to our strategic planning. It will allow us to better meet our customer needs. For example, for our CSP customers, we can give them dedicated production lines. This will help us better meet our customers' orders. In 2027, I cannot yet give you full visibility, but I can be quite confident that the second half will be better.
Good afternoon. I have many questions, so maybe I should ask them one by one. First of all, you mentioned that in the second quarter, one of your major customers had a delay in their shipments into the second quarter. Can I ask, what is their contribution to the second quarter revenue?
Well, it's our biggest customer, so it's a pretty big share of our revenue.
Is it for a specific project that was supposed to ship in the first quarter, but because of various factors, it made a big order for the second quarter? Okay, understood. My second question, for IT & Consumer Electronics, you mentioned that there were orders pulled forward to the second quarter. You mentioned concerns about a slowdown in the second half. Has this changed?
No, this has not changed, and it may become even worse than our previous forecasts. Yes, it may not even reach the level of our previous predictions. I mean the market as a whole.
What about for your company?
Well, for our company, the impact would not be so large because when we talk about market demand, we're talking about volume. For us, we focus on pricing and revenues. Because of pricing increases, we can blunt the impact. For example, one of our customers, they lost 20% of their market, but their revenues remain mostly unaffected because they were able to increase their price by 20%.
Okay. For the third quarter, for the ICT segment, will it continue to grow quarter-over-quarter?
Well, I wouldn't say that it will grow, but I will say that because of price adjustments, we can at least hold steady.
Okay. My next question is about the TWD 18 billion CapEx. Can you break that down into equipment and land?
Well, to answer this, of this TWD 18 billion, about 30% will be in land and building expansions. The remaining 70% will go to equipment, technical R&D, automation, et cetera. That would be 70%.
All right. To follow up on that, will there be an increase in depreciation this year?
No, I do not anticipate this. When it comes to the buildings, the depreciation is relatively slow, and for equipment, I don't see much of an effect this year either.
Okay. You mentioned two new CSP customers giving orders this year, and you also said that the target for each key account is $10 billion. What is the timeline for achieving this target?
Well, this is how I see it. If the next generation product is introduced smoothly this year and mass production is achieved next year, we can hit this target in two years.
Okay, so if you've started supplying power supplies for these two customers, is it for ASIC or GPU?
Well, both. Both GPU and ASIC.
Okay, final question. You talk about cloud, and you say that it could grow by 70% this year, but it seems that for the first half of the year, it grew by over 100%. Yes, the first half of the year saw a doubling. What about the year as a whole? Will this continue 100% growth, or will it be more like the 70% that you talked about?
Well, I think we will be confident in maintaining at least 70% growth.
Can you expand a little bit on that? If the cloud segment grows by 70% this year, what about the BBU year-over-year? What about the BBU segment specifically?
Well, if you look at it this year, BBU growth will be 20%-25%. I mean, as a share of the cloud business as a whole. If we're talking about growth momentum, it's about the same as other product segments. Yes. Mostly it's because the production capacity is coming online.
In the past, the BBU segment has not been less than 20% of the cloud segment as a whole, right?
Yes, because our production capacity is growing, when we look at the year as a whole, we believe that BBU will account for close to 20% of the cloud segment as a whole. Our guidance at the start of the year was that the cloud segment will grow by 70%, and we believe that the BBU growth will be in line with that. I was wondering if the BBU would grow faster than the other products.
It would depend on production capacity, is what I can say. I think it's a trend as a whole that in the future HVDC market, BBU will become a standard part of that. We believe that it will continue to grow.
Can I confirm once again that for the BBU products, it will see even better growth in the second half of the year? Is it the highest growing product in your cloud segment?
I don't think it's the highest growing single product because BBUs are part of a unit with PSUs and Power Shelves, et cetera. I think when it comes to unit pricing, the Power Shelf is the highest. As Julia said, BBU accounts for only 20%-25% of the cloud segment as a whole.
To ask once again about DenseLight. It's a startup, because I see that it's still seeing relatively low revenues, and we are investing over 18 times the PS ratio. Why? What do you see in that company?
Well, I think our investment has several reasons behind it. First of all, it's an IDM company. It doesn't have just process, it also has production as well. It's a whole package, and it's done all in one site. That would be very important for future iterations of optical communications technology because it allows us to maximize the chips, and with our optical coupling and packaging technologies, it would be very beneficial because light loss is something that we all really care about. As an IDM company, that's one major reason. The second major reason is that it is a purely Singaporean company.
All of its production and R&D are in Singapore, it is part of the non-Chinese supply chain. That's the second question. Second reason rather. Third, it has been focused on optical communications for 20 years now, it's been committed to this segment for a very long time. It was because of the development of data centers that's pushed them to work with us because there are now more opportunities for them. Those are the three reasons that we're embarking on this strategic partnership.
Okay, we can accept one last question.
Hello, managers. I want to ask you about DC-DC. It sounds like that DC-DC is a product that you are deploying and developing recently, including your buying of the Nanjing company. For IBC or for the next level, for the future platform, Power Rack, I think Power Rack and IT Rack are separate. This should be put into IT Rack. Under the new platform, does customers ask you to ship in bundle, or do you simply want to win this market? I'm not pretty sure why.
Let me put it this way. In the future, HVDC from 800V to 50V to 48V or 12V, the applications are already there's no question of bundling. This is essential need on the market. From the perspective of energy management, using DC allows for better energy efficiency. In the future, it will go down from 800 directly to 6 V. The technology threshold is going up. This is something that we want. This is our perspective. The DC-DC market has always existed, but we didn't participate much because in the low-end market, margin wasn't high.
When you deliver Power Rack to specific customers, there's no DC-DC bundling, and you have to further win this market, right?
Yeah, this is another thing, another service or value that we can provide to customers. In fact, this is why they expect Lite-On to enter this market to provide this product for them.
Is it related to technology or is it not related to platform?
Well, it's highly related to technology.
When you deliver Power Racks and when they use your DC-DC, can the efficiency go up?
Well, you can look at them as two different components. Power Rack is one thing and DC brick is another thing. Thank you.
Okay. This concludes our earnings conference today. The video and all the materials will be put onto our official website. Thank you so much for your participation. Thanks