MediaTek Inc. (2454) Earnings Call Transcript & Summary

Gentlemen, welcome to the MediaTek 2024 First Quarter Investors Conference Call. Financial results and presentations for today's call are available on the Investors section of the company website at www.mediatek.com. And now I would like to turn the call over to Ms. Jessie Wang, Deputy Director of Investor Relations. Ms. Wang, please proceed.

Good afternoon, everyone. Joining us today are Dr. Rick Tsai, MediaTek's CEO; and Mr. David Ku, MediaTek's CFO. Mr. Ku will report our first quarter results, and then, Dr. Tsai will provide our prepared remarks. After that, we will open for Q&A. As a reminder, today's presentation will provide forward-looking statements based on our current expectations. The statements are subject to various risks and factors which may cause actual results materially different from the statements. The presentation materials supplement non-TIFRS financial measures. Earnings distribution will be made in accordance with financial statements based on TIFRS. For details, please refer to the safe harbor statement in our presentation slides. In addition, all contents provided in this teleconference are for your reference only, not intended for investment advice. Neither MediaTek nor any of independent providers is responsible for your any actions taken in reliance on contents provided in today's call. Now I would like to turn the call to our CFO, Mr. David Ku, for the first quarter financial results.

Thank you, Jessie, and good afternoon, everyone. Now let's start with the 2024 first quarter financial results. The currency used here is NT dollar. Revenue for the quarter was TWD 133.5 billion, up 3% sequentially and up 39.5% year-over-year. Gross margin for the quarter was 52.4%, up 4.1 percentage points from the previous quarter and up 4.4 percentage points from the year ago quarter. This quarter's gross margin include a one-time cost reversal item for securing capacity in the past, which increased gross margin by 4.5 percentage points in this quarter. Excluding this one-time adjustment, gross margin was 47.9%. Operating expense for the quarter was TWD 37.7 billion compared with TWD 37.9 billion in the previous quarter and TWD 31.5 billion in the year ago quarter. Operating income for the quarter was TWD 32.2 billion, up 30.1 percentage sequentially and up 124% year-over-year. Non-TIFRS operating income for the quarter was TWD 32.4 billion. Operating margin for the quarter was 24.1%, up 5 percentage points than the previous quarter and up 9.1 percentage points year-over-year. Non-TIFRS operating margin for the quarter was 24.3%. Net income for the quarter was TWD 31.7 billion, up 23.1% sequentially and up 87.4% year-over-year. Non-TIFRS net income for the quarter was TWD 31.9 billion. Net profit margin for the quarter was 23.7%, increased 3.9 percentage points from the previous quarter and increased 6 percentage points year-over-year. Non-TIFRS net profit margin for the quarter was 23.9%. EPS for the quarter was TWD 19.85, up from TWD 16.15 in the previous quarter and up from TWD 10.64 in the year ago quarter. Non-TIFRS EPS for the quarter was TWD 19.98. A reconciliation table for our TIFRS and a non-TIFRS financial measure is attached in our press release for your information. And that concludes my comments. Thank you.

Thank you, David. Now I would like to turn the call to CEO, Dr. Rick Tsai for prepared remarks.

Good afternoon, everyone. MediaTek delivered solid first quarter. Revenue exceeded the high end of our guidance, mainly driven by better-than-expected restocking demand from smartphone, broadband and TV customers. Gross margin exceeded guidance due to a onetime item, excluding which gross margin was 47.9%, above the midpoint of our guidance range. In the last earnings call we estimated overall consumer demand to improve moderately in 2024. Our view remains unchanged today. Given better visibility than in the third -- in the last quarter, we now target our full-year revenue in U.S. dollars term to grow by mid-teens percent in 2024. As for full year gross margin, excluding the first quarter one-time item, our target remains to be 47%, plus or minus 1 percentage point. While all revenue groups are expected to grow year-over-year, smartphone growth is expected to be stronger compared to other revenue groups. Specifically for smartphone, global shipment in 2024 is expected to increase by low-single-digit percent to 1.2 billion units. Our addressable market in dollar terms can grow higher at low teens percent, driven by continuous 4G to 5G migration and product mix shifting towards higher-end phones. Therefore, for MediaTek, coupled with our strong market share gains and higher blended ASP in the flagship segment, we are confident that we can grow our smartphone revenue by higher than mid-teens percent in 2024. Flagship revenue is expected to grow more than 50%. For the mid to long term, AI represents great business opportunities for MediaTek, both in the cloud and at the edge. In the cloud, with the proliferation of AI applications, we see increasing AI accelerator demand from cloud service providers to further strengthen their AI computing. MediaTek's strength in computing technologies, coupled with the leading 112G and 224G SerDes IPs are crucial for winning AI accelerated businesses. In addition, we are equipped with strong capabilities in complex IC integration, advanced process nodes and advanced packaging, as well as a flexible ASICs business model to grasp future opportunities. At the end -- at the edge, in addition to our leading APU solutions, MediaTek's NeuroPilot platform provides comprehensive software tools supporting all the popular AI models. These include Meta's latest Llama 3, Google's Gemini Nano, Stable Diffusion models and all Mainland China's popular generative AI models. MediaTek's NeuroPilot is well received by the developer community and creates a strong ecosystem to accelerate edge AI development. With that, now let me talk about the recent business performance of our 3 revenue groups. Mobile phone accounted for 61% of total revenue in the first quarter, growing 84% year-over-year and declining 2% sequentially. As stated earlier, our mobile phone business will benefit from a better product mix this year. And we're excited about the opportunities generative AI will bring us. Smartphones are viewed as the most imminent generative AI edge device for content creation and personal assistant related features. Our powerful Dimensity SoCs, together with our growing generative AI ecosystem, including our NeuroPilot platform are ideally suited to empower smartphones with richer AI functions. Today, global top 3 smartphones ranked by AI performance are all powered by Dimensity 9300. There will be more Dimensity 9300 and 8300 smartphones coming to the market in the second quarter. Our next generation flagship SoC, Dimensity 9400, is on track for launch in the second half of this year and has received a very positive feedback from customers. Now let me move on to Smart Edge platform. This group grew 16% sequentially in the first quarter and accounted for 34% of total revenue, mainly driven by restocking demand from broadband and TV customers, as well as product mix enhancement in tablet. We continue to see upgrade demand for connectivity and computing driving growth for this year. For connectivity, in the first quarter, our wireless and wired solutions benefited from our previous operator's adoption of the latest technology such as Wi-Fi 7 and 10G PON. In particular, our leading Wi-Fi 7 solutions are tracking ahead of our prior expectations and ramping robustly across broadband, routers and notebooks. For computing, tablets powered by Dimensity 9300, which supports generative AI, began mass production in the first quarter, and we expect more adoptions to come. For automotive, our Dimensity Auto Cockpit solution continued to secure new design wins with leading carmakers for future revenues. We're also working with NVIDIA to bring powerful generative AI capabilities to vehicles across premium and entry-level segments with our Dimensity Auto Cockpit chipset. Now moving on to Power IC, which accounted for 5% of total revenue in the first quarter and declined 13% sequentially, mainly due to seasonality and product transitions. For the second quarter, we expect revenue to grow strongly year-over-year. On a sequential basis, revenue is expected to range between flat and 9% decline as smartphone shipment reaches a normal pattern in the second quarter. Meanwhile, TV and computing devices are expected to continue to grow sequentially, and power IC demand to recover across applications in the second quarter. With that, we expect our second quarter revenue to be in the range of TWD 121.4 billion to TWD 133.5 billion, flat to declined 9% sequentially and up 19% to 30% year-over-year at forecasted exchange rate of TWD 32 to USD 1. Gross margin is forecasted at 47%, plus or minus 1.5 percentage points, in line with the gross margin in the previous quarters. Quarterly operating expenses ratio to be at 30%, plus or minus 2 percentage points. With the first quarter results and the second quarter guidance, we are confident of achieving our annual revenue growth and gross margin targets. Our view for the mid- to long-term growth outlook remains intact. AI is driving huge business opportunities both in the cloud and the edge for MediaTek. Our key design wins and new project executions are progressing well to contribute revenues from the second half of 2025. We believe 2024 will be the beginning of our next growth phase. Thank you.

Thank you, Rick. Operator, we are now ready for Q&A. We're pleased to have the first question.

[Operator Instructions] The first one to ask questions, please, Laura Chen from Citigroup.

And I appreciate Rick's very comprehensive description about MediaTek's outlook. I'm just curious about that when we move to the AI arenas, like for the AI smartphone or edge AI, aside from the large size of the chip for neural engine, what's MediaTek's observation of the technology upgrade at the hardware architecture such as like thermal design, memory, et cetera? And also, what MediaTek is currently doing for better power consumption or heat dissipation, et cetera? And for instance, like do you think that current PoP type packaging needs to be upgraded? Or anything you are doing? Can you share with us? That's my first question.

Well, Laura, your question maybe -- can be divided into 2 areas, one for cloud and the other for the edge. For the cloud, it is obvious that the key -- I think, some of the key technology areas include, of course, the leading-edge process nodes, today at 3-nanometer, soon into 2-nanometer. Advanced packaging technologies, the 2.5D, may be 3D, provided by mainly a foundry, key foundry suppliers. Memory, the high-bandwidth memory certainly is now getting to be a critical component and sometimes maybe a bottleneck component for the data center AI deployment, which is moving from, I think, HBM3, HBM3e and very fast into HBM4. And lastly, all the major interconnect -- high speed interconnect IPs such as SerDes -- high speed SerDes, up to 224G is also key because interconnect with all the computing that you need to put, you need to connect every socket or -- no, every server, every rack. The interconnect technology's IP is becoming another bottleneck. So while we thought -- and if we look at MediaTek's capabilities, and we are well-suited to move into the AI accelerator, mainly in the compute part. We are now really also targeting and focusing in that area. For the edge devices, I think as we stated in our opening remarks, our Dimensity 9300, and I also -- I'm very confident about our 9400 to be a leading SoC solution for the edge generative AI applications. We are -- I think our AI capability within our SoC is at least, I think, as good as any. And we -- again, we believe this will continue to drive our smartphone business together also into some other computing devices and automotive applications. Thank you.

Just following that, for the edge side, other than the neural engine becoming bigger and more advanced now, do you also see that on the packaging side, there is a major change looking forward for the edge device?

Not particularly. I think we are moving -- no -- I mean, I think depending on your definition of edge, for the smartphone, I think not too much different. But as we also move into automotive applications in the cockpit, and later on, likely into the ADAS application, the packaging technology will move into more advanced sophisticated one. But more advanced than smartphone, but less compared to the data center AI processes. Thank you.

A follow-up question may I have on the gross margin impact because we know where gross margins have been quite stable. And with those new technology and probably the rising cost of the foundry and also moving to more advanced packaging potentially as well, so how should we look at our profitabilities in the longer term? Is that 47% or 48% gross margin, may see further upside if we move to more advanced technology?

For now, we -- I mean, as we -- again as stated earlier on, we are confident that we can achieve 47% plus/minus going forward. We remain convinced of that. We understand your question. I think the key here is whether we can make products with more higher value, more and higher technology contents. I think that there will be some cost impact from advanced -- the leading edge process node and say 2.5D packaging. But we are also making products using those technologies with much higher content, technology content and values. So if you look at the gross margin and the operating margin, I'm quite comfortable that we will maintain, manage our margins hopefully better going forward. Thank you.

Next one to ask questions, Randy Abrams from UBS.

I wanted to ask actually a question on the outlook for second quarter. I think you referred to it as a normal decline. I'm curious if that's the -- if you think it's the new normal pattern just given the flagship launches. Traditionally you saw a rebound or a seasonal ramp in second quarter. And just how you're viewing the broader Android smartphone cycle, if you're viewing some inventory adjustment after the strong restocking? And if you think it's a 1 quarter issue that third quarter we're back to growing in the smartphone business?

Well, Randy, I think if you look at the overall market, we said it's 1.2 billion units, which is about maybe 2% increase year-over-year. It's not really a major increase. And so at the end of the day, every quarter -- at the -- after 4 quarters, we will balance out. There's no question that the fourth quarter last year and first quarter this year represent a pretty strong restocking demand and as a result our sell-in to our OEM customers. But this trend is -- as I said, is now reaching a normal level. So that's why we're seeing some impact from a sequential point of view. Saying all that, we also said in our remarks that while the overall demand increased by low single digit in the unit, but the high-end segment -- by high end, we can broadly define as I'm going to use the RMB pricing, RMB 4,000 and up model. I think the growth in that area -- in that segment is much better compared to the other segments. And as we also are making progress and gaining shares in the flagship SoCs. So our addressable market is quite a bit higher in the teens. So I'm okay. I think for the whole year we give you the outlook also for the mobile phone. We are very confident in that.

I'll ask a follow-up to that. There's been a lot of remarks this results even it's relatively muted first half, but some are expecting we start to see a bit of pick up second half. I'm curious, when you look across your businesses, how you see smart edge and power management? How you're viewing kind of inventory levels demand, like the initial cut what you're seeing, if you're seeing improving prospects from those as we look out half-on-half?

I think, Randy, we have given outlook in U.S. dollar terms for our yearly revenue growth while being a target of course. And you have now the first half numbers. But with the guidance, I'm sure you will have the second half estimate. Again, we -- at the beginning of the year, we stated that the 2024 is a moderate growth year. Overall, we remain that. And so the second half number is not surprising. The restocking momentum actually demonstrates the most in the smartphone products. Other products, I think, more stable. There were some in the first quarter for the TV and the connectivity, but the magnitude of which is more moderate in nature. So, again, we said what we said, being a moderate year. And I believe the number we give, the mid-teens growth for the whole year, it's a very, very achievable target. Thank you.

Okay. Yes. And the last question I wanted to ask on the flagship where you mentioned the 50 -- over 50% growth. Is that mainly from the gains you see current cycle? I'm curious as you go into the new cycle, the early design activity and as you look at the different pieces, Qualcomm will move to the NUVIA core, so they upgrade the CPU. But how are you looking? Are you viewing the system performance with more attention now on the AI engine in that software? Like how do you view your growth moving into the next cycle at this early stage? And also, how do you view content, where you're -- it sounds like you're going to upgrade the AI engine and also the higher cost from 3-nanometer? How you see the content for the flagship category?

Okay. For the flagship SoC, 9300 is our third-generation. SoC, 9400, of course, the fourth one. I think that we started 9300 as kind of a -- I wouldn't call that breakthrough, but it's a major leap in our -- not only the capability, but also our penetration with the key customers. And we are seeing a very strong growth from 9300. And I have full confidence in our 9400, especially in our computing capabilities. The CPU core that we're working with our partner -- our close and deep partner, we have no issue with our performance and our power capability for our 9400 CPU cores. And for the APU modules -- did we say our TOPS before?

Not yet.

No. Not yet. Okay. I guess I'm not saying. I think I'm not supposed to. Yes. But I can guarantee you, 9300 were about 37, 35 TOPS this year and a significant improvement for 9400. And this -- actually, I remember I talked to one key customer, and they are surprised to hear such data from our flagship SoC. With that and with our NeuroPilot platform, we are now also engaging with most of the developer community and the ecosystem players. I think the key here is to enable all the applications with the edge devices so people can find not only edge AI being used for productive and, I believe, more importantly fun for their life. And there's a great future for this flagship SoC business for MediaTek.

And could you say the content -- and then I'll drop back in the queue, but could you say the content increase factoring those enhancements and the node migration?

Yes. I think I kind of say that it's mainly especially in our computing, the CPUs, the GPUs and the APUs and of course ISPs. They are getting -- we are using up the -- all the 3-nanometer -- well, of course, there's -- there's a limit to the die size for the SoC for phone format. But the CPU contents and memory -- I'm sorry, computing content continue to increase and the -- but you have to also realize our computing contents increase while our power consumption capability improves. I think that's the -- that's what we've been able to do and that's what we are -- get ourselves more market share.

The next one is Brett Simpson from Arete.

Rick, I wanted to ask about just the AI smartphone market and what portion of the overall smartphone market might be enabling Gen AI this year? And how do you see the adoption curve playing out into the second half and maybe into next year? And what sort of market share of the Android AI smartphone market does MediaTek hope to achieve over the sort of medium to long term?

Well, Brett, I think just your last question, I think we can achieve 30% with China the Android smartphone -- flagship smartphone market share short term. Hopefully, of course, we will aim higher. I think many people asked a question about whether Generative AI will accelerate the replacement cycle for instance. What I believe and I observe, maybe can be set in 2 aspects, one, the applications from the GenAI, I think, it's now in the early stage of being deployed, and well, maybe also spread into the user environment. The fact that the move towards the high end phone, as by the way, this movement is pretty -- very fast. So the high end phones definitely is growing very fast. I think kind of exemplifying that, the trend as more people get more used to or maybe addicted to GenAI applications, I think the trend can only accelerate. So, of course, along with the much higher content, the technology, silicon content. So again, for us, as I said just now, another critical thing is to accelerate the application development and deployment together with our partners in the ecosystem and the developer community. Only in doing so can we together expand the market and have consumers enjoy the benefits. Thank you.

And Rick, just to follow up on that. You talked about the TOPS performance of the 9300. I think we're all expecting a big upgrade with the 9400 later this year. But can you maybe just talk a little bit about how this translates to content growth for MediaTek in smartphones or the die size that's allocated to the NPU? How it changes? I think most folks are expecting maybe about 60 TOPs in the next-generation chips that are coming out for flagship later this year. But how does that -- how do we think about the content translation for MediaTek?

Actually, Brett, you certainly have done your homework. I'm not going to argue again to your comments. We -- I think 9400, the -- our CPU core, I really feel very strongly in a good way, so the clock rate, the performance of the IPC and not to mention our APU's TOPS. I really feel 9400 will make 9300 -- well, really outshine 9300 by quite a bit, by quite a bit. I know I'm being kind of a qualitative, but I think I'm conveying a very clear message in this way. Thank you. The die size, by the way -- Brett, die size, no matter what technology we use, 3-nanometer, we still have 1 -- the die size, we use supporter die size, period, and a lot of which goes to computing, as you can imagine.

Yes. For my second question, Rick, I wanted to ask about Wi-Fi 7. And just to understand the ASP change that we're seeing here. I think MediaTek is making a double shrink. I think you were on 22-nanometer for the previous Wi-Fi generation, and you're moving to 6-nanometer with Wi-Fi 7. And you're definitely viewed as among the Tier 1s in this transition. So can you maybe talk a little bit about your market share ambitions and what sort of ASP changes we see as Wi-Fi 7 becomes more mainstream?

Brett, David here. I think Wi-Fi 7 is overall, from the size of market and also from technology segmentation perspective, we're actually one of the leaders in the market. And because of that, it really just opens a lot of new opportunity for us. But in terms of the absolute market share, we probably won't be able to disclose right now. But overall, we will say that is pretty meaningful market share for the Wi-Fi 7 given that is our earliest access to the technology and also actually the breadth of product portfolio.

Next one to ask questions is Gokul Hariharan from JPMorgan.

My first question is on the smartphone side of things. So last call I think, Rick, you mentioned the flagship addressable market TAM is about 4 billion. Could we talk a little bit about how quickly is this TAM growing given that you highlighted the strong growth that we are seeing in the flagship segment? And could you also comment about any prospects to expand this TAM by kind of breaking into some newer larger customers beyond the customers that you already shipped to? Can we expect something happening in that front in the next 1 year or -- 1 to 2 years or so?

We certainly. We cannot really discuss specific customers per se, but we certainly -- again, I think our product, in my firm belief, will sell itself because of our capability. And we also see the OEM customers now, of course, using the last generation flagship for their kind of a sub-flagship phones. All those increased the addressable market. So the driving force, really underlying driving force is the move towards the high end phones, be they the flagship or the, what we call, premium phones. The trend really provides the addressable market driving force for us. Thank you.

So any thoughts on how quickly this is growing? I think given that last quarter you offered a 4 billion TAM number for your immediate flagship TAM.

Yes, I think we talked about 4 billion being our TAM -- MediaTek TAM. Again, I think that addressable market, as I said earlier, will see a mid-teen percent level of growth. And that's why we are stating that we -- that part of the reason that we are stating our growth of 50% and better in the flagship SoC area. Am I answering your question?

Yes, I think that's clear. My second question is on the data center of ASIC. I just wanted to understand how MediaTek is positioning itself in the data center ASIC market with especially the CSP customers. I think some of your competitors, I think Broadcom, obviously, has talked about pretty high market share. Marvell, I think, recently talked about having -- or targeting 25% market share in custom ASICs for accelerators and other products by 2028. So just wanted to understand how is MediaTek positioning? When you do your business planning, what are the kind of targets that you're looking for? I think, if I remember, a few years back, Rick, you talked about potentially hitting a 1 billion run rate for the non-consumer ASIC business. How soon can we anticipate something like that happening for MediaTek?

Yes, we definitely are on our way to an ARM computing revenue stream in the near future. But in addition, again, AI data center, what we call data center AI accelerator and mainly in the compute area, represents a very large TAM, certainly for MediaTek also. We look at that TAM to be about $40 billion by 2028 time. It's about $12 billion in 2024. So it's a tremendous upside, and the -- and as I said earlier, the key component to get to this compute AI accelerator business, computing, obviously, the leading-edge process node, advanced packaging technology, those foundation technology. And please also remember to connect the cloud and the edge, we have all the wireless and wired connectivity technology. And also the way we are now working with ecosystem players, I really believe that we will position MediaTek to be one of the major players in this -- well, by 2028, $40 billion TAM business. We are -- the company is certainly investing heavy in this.

So, Rick, maybe one follow-up is, of the $40 billion TAM, what would you consider a reasonable success? Is it like 10% market share by 2028 would be a reasonable success for MediaTek?

Well, I don't think we will give this forecast yet. But, yes, I think that kind of number is certainly within our expectations.

Next one to ask questions, Bruce Lu from Goldman Sachs.

I want to follow up with the last question earlier. What does the AI smartphone penetration looks like for 2024 and 2025? And moving beyond the edge or inference, what is the latest strategy for MediaTek? What kind of business model and value proposition for your MediaTek DaVinci? What kind of business do you want to do for that product? What is your expectation for the AI edge?

Maybe commenting on DaVinci first, and also, for the AI, the smartphone penetration, probably Rick can talk about that. I think DaVinci essentially is a platform -- is a co-work platform MediaTek uses internally. Basically, that's one of the effective tool for us, for all our engineers, just a tool using the GAI tools because it's actually very powerful and also very useful. A lots of the external parties are trying to match that. So we do actually open up to other parts, but to be honest, this is actually not a big business model. This is more for the operations side. So I would say probably we don't need to mix that with our business outlook. It will be easier. So that's about DaVinci.

Bruce, the way to look at, what you just said, AI smartphone penetration, I believe actually the best way right now is to look at the -- again, the high-end smartphone's volume and market share, how that has grown this year and how that will continue to grow in next year. It's very difficult to get the numbers. This is the AI smartphone or AIPC for that matter because what we know and what we are working very hard on is to ensure that our chips can provide all the technology capability and to enable all the Generative AI applications and future acquisitions for the OEM customers. And we are fully confident in that. And this, for the AI, Gen AI application, to be developed and deployed, takes of course an industry-wide average, ecosystem-wide effort. MediaTek actually is taking a pretty proactive attitude in that we will have some major events in the coming weeks from that point of view.

But I thought you got to proliferate into the mid-end chips next year. So most of your mid-end chips will have the AI function that could potentially be a big boost for the AI smartphone penetration rate? Is that the right expectation?

Yes, I think it's the right expectation. I think, firstly, for this year's, so called, APU is already 9000 series and also 8000 series as well. And so like Rick said, right now, the key objective is really trying to proliferate about the AI ecosystem because for the hardware side, we do have the leading APU there, but now I think we need more ecosystem partner to jump on the wagon to basically develop more applications or service, both on edge or leveraging edge device. But going forward, we do have the plan to extend that APU capability even into the different segments, into different segments.

Okay. My next question is for the ASIC business. So do you want to do the pure design service business with lower gross margin but a huge operating leverage? Or do you want to provide more IP value? For example, Rick just mentioned a lot about like 2 to 4 giga SerDes, which is a key successful factor for AI accelerators. So what is our latest progress? And how confident that we can be very competitive in this -- in all the IP required for the ASIC business?

I want to say that we definitely will be and want to be an IP-rich value, high value-add supplier. Well, of course, we always remain flexible in our business model. The thing really is that we -- the importance is we invest in those high value-add IPs. Well, SerDes just being a good example, not to mention the process node, the packaging and the -- I think we, MediaTek, is -- I mean, we strive to be a Tier 1 supplier. Maybe that's another way of saying it. Thank you.

[Operator Instructions] And now it's time for Brad Lin from Bank of America.

I have 2 questions. One is on the -- well, regarding the ASIC. Regarding this industry landscape, besides the, well, current players, we also see, well, many more so-called established large fabless company began the investment. So would you mind and elaborate the strength that MediaTek has compared to them? And also, what will be the key difference whether a firm joins the ARM Neoverse or not? That's my first question.

Okay, Brad, I think it's like Rick reiterated several times. I think, first of all, the focus of what you are talking about on the cloud or on data center side is really just the AI accelerator and especially for the compute, especially on the computer because actually there are other components out there as well, but compute is the key. And transitionally, I think that's including CPU, GPU, and now, we're talking about XPU, most likely for TPU. I think with our technology and also especially the foundation technology like SerDes, advanced SoC capability and also the leading-edge process node technology. I think we are actually one of the key players and important player to compete effectively. Well, another part actually is the ecosystem partner and also our actually the operation scale. Because that all, by end of the day, it's actually we have the right cost structures, both from an operation side and also from the wafer side. I think that could be a very important element as well to stand us out among all competitors, especially for the leading those parts. So I think -- all in all, I think the key word, which is the computing technology, and very much, MediaTek product is one of the leading companies to do so many connected computing devices every year. And we will just extend that technology and also advantage from the traditional edge device, now actually into the cloud computing element. Likewise, all the foundation technology, a lot of that will be leveraged and expand into the cloud space as well.

Got it. So my second question will be on the PMIC. While we know MediaTek has done quite well at bundle selling all the PMIC offerings, what will be the key growth drivers into the next 2 years?

Okay. So, PMIC, the areas that we are really focusing in and invested -- investing in is in automotive and the data center area. We actually -- we have formed an R&D team in the U.S., in Japan so that we can build those leading -- well, I would say, leading-edge PMIC. We also invest in the process capability with our foundry partners so that we can take advantage of the light power devices to gain advantages. So I think we have good effort, and yes, we expect good growth from our PMIC business. Thank you.

So last thing on dividend. MediaTek shareholders obviously have received special dividend in the past 3 years on top of the 80% to 85% payout. Should we expect MediaTek to start another round of special dividend or raise the payout ratio?

I think currently there is 2 elements for our tax system. One is the payout, another one is the special dividend. I think for payout, basically, nothing's changed right now. For the special dividend, we're still actually on a very last payment of the special dividend. For the new one, we will discuss with the Board and also report everyone when it's actually is -- once finalized.

The last one to ask questions, Charlie Chan from Morgan Stanley.

So again, still congrats for very good results, margin side, and also, 2Q revenue full year seems to be going well. So my first question is about -- Rick, you seemed to suggest that high-end smartphone demand is doing well. I'm not sure if you relate it to the AI. I'm wondering whether MediaTek has any evidence or feedback that the high-end smartphone demand is related to the AI. And if there is a kind of a so-called killer apps, what would that be?

Again, Charlie, I think I said that earlier. Very early stage for the Generative AI applications, actually any devices if you think about it. The people need, I believe, to get their devices which are capable of exercising those applications when they become available and when they become also attractive. And that's hard, and I think that's why we are taking a very proactive stand in working with the ecosystem and the developer community. So we can together enable new applications and exciting applications. So my view for this is, of course, it's also very -- just like if you ask whether -- how many people are buying the camera capability for their phone, very difficult to quantify. But the important thing for us is not to try to cherry pick, okay? We provide the -- just a very powerful phone and very low power consumption to our consumers.

I see. Yes. So, yes. Yes, because the company seems to continue to enlarge the APU capability. So I thought you might have some reason behind, right? For example, you already see the killer apps for the next generation apps, yes.

Well, first thing first, we have to be able to, for instance, run 7 billion parameter model or up to 13 billion model. And we need that kind of computing power. But those are, I think -- again, are the must for our flagship chips.

Got you. And my second question is about the smartphone cyclicality. So I understand that we are kind of in a destocking mode for China smartphone. First of all, how long do you think this starting is going to last? I'm not sure your view about China economy's recovery. And because the company gave the full-year revenue guidance, right, so I'm surprised that you already have visibility for the second half. Are you confident about the second half recovery?

Charlie, it's David here. First of all, I think seasonality, if you look back in the last few years, including this year, the so-called seasonality may be very year-over-year. Taking last year, for example, we're ramping from first quarter to fourth quarter, we see very strong quarter-over-quarter growth. And that momentum left until first quarter this year. And -- but the second quarter for us has become mathematic. We have a high first quarter, second quarter coming down. But if you look, the key issue is looking for the full year, that's why I think Lih could actually spend some time in the opening remark talking about we always start with the year-over-year that sell-out, which is low by the way, 2%, okay? But you build up from that, and that gives you the product migration, product segmentation, and that actually leads to our mid-teens. So that's the answer to you directly. Our confidence is not based on so called strong rebound with the confidence really based on year-over-year only 2% in shipment growth, okay, around the increase. And truly confidence is really just the product segmentation expansion and more importantly a very successful smartphone expansion and also growth. And with that smartphone, we tend to know what's the design situation right now, and also, even for the fourth generation, 9400, we're trying to know the designing situation already. So we feel fairly confident for our growth topic. I think that's it. But anything else in between is really just a patter. What really matter is the full year's dollar growth. I think that's the key.

One more word, Charlie. The smartphone sales goes out from China also to overseas market. China market is a big part, but it's not the only part. And I understand your point about China maybe not growing much as a whole. Well, we certainly don't disagree. However, the China market's, again, that movement towards the high-end phone is significant. It is ongoing, and you will not -- we don't believe that will stop. However, the market outside of China is partly compensating for some of China's lot -- I mean, not buying, for instance, the entry-level phone as many as before. Other people, the 4G to 5G migration is continuing, and actually, if anything, is probably accelerating at the low-end 5G segment. So all things together, as David just said, if you look at the whole year, I think we are -- we cannot say we know exactly what will happen, but we're quite confident. Thank you.

Ladies and gentlemen, we thank you for all your questions. Now I'll hand it over to Ms. Jessie Wang for closing comments. Ms. Wang, please go ahead.

Ladies and gentlemen, this concludes MediaTek 2024 First Quarter Conference Call. And an audio replay will be available in 1 hour after the call at the Investors section of MediaTek's website. We would like to thank you for your participation, and you may now disconnect. Thank you.

Thank you, Jessie, and thank you, ladies and gentlemen, for your participation in today's conference. You may disconnect your lines now.