Silicon Motion Technology CORP Q1 FY2021 Earnings Call

Good day, and thank you for joining us. Welcome to Silicon Motion Technology Corporation's First Quarter of 2021 Earnings Conference Call. This call includes forward-looking statements as defined by the Securities Act of 1933 and the Securities Exchange Act of 1934. These forward-looking statements cover trends in the semiconductor industry and our anticipated operational results, financial condition, and business outlook. While these statements are based on our own information and reliable external sources, they should not be overly relied upon. There are risks and uncertainties that could cause actual results to differ significantly from those projected. Key risks include ongoing competitive pressures in the semiconductor industry affecting prices, unpredictable technological changes and consumer demand for multimedia electronics, the nature of our relationships with major customers, and shifts in political, economic, legal, and social conditions in Taiwan. For more details on these risks and uncertainties, please refer to the documents we periodically file with the Securities and Exchange Commission. We are not obligated to update any forward-looking statements, which reflect our position only as of today. Also, please note that today's conference is being recorded. I would now like to introduce your first speaker for today, Mr. Chris Chaney, Director of Investor Relations and Strategy. Please go ahead.

Thank you, Annie. Good morning, everyone, and welcome to Silicon Motion's First Quarter 2021 Financial Results Conference Call and Webcast. As Annie mentioned, my name is Chris Chaney. I'm the Director of Investor Relations. Joining me today on this call are Wallace Kou, our President and CEO; and Riyadh Lai, our Chief Financial Officer. Following my comments, Wallace will provide a review of our key business developments, and then Riyadh will discuss our first quarter results and our outlook. We will then conclude with a question-and-answer period. Before we get started, I'd like to remind you of our safe harbor policy, which was read at the start of this call. For a comprehensive overview of the risks and uncertainties involved in investing in our securities, please refer to the filings with the U.S. Securities and Exchange Commission. For more details on our financial results, please refer to our press release, which was filed on Form 6-K after the close of the market yesterday. This webcast will be available for replay in the Investor Relations section of our website for a limited time. To enhance investors' understanding of our ongoing economic performance, we will discuss non-GAAP information during this call. We use non-GAAP financial measures internally to evaluate and manage our operations. We have, therefore, chosen to provide this information to enable you to perform comparisons of our operating results in a manner similar to how we analyze our own operating results. The reconciliation of our GAAP to non-GAAP financial data can be found in our earnings release issued yesterday. We ask that you review it in conjunction with this call. And with that, I'd like to turn the call over to Wallace.

Thank you, Chris. Hello, everyone, and thank you for joining us today. In the first quarter, we delivered record sales and earnings per share. Sales grew 27% sequentially to a record $182 million, and earnings per ADS were a record $1.11. Our strong revenue growth was driven by strong sales of both our SSD controllers and eMMC+UFS controllers, with both products achieving record quarterly highs. The outlook for the upcoming quarter and the rest of the year continues to be very strong. And as we have secured additional wafers from our foundries, we are raising our full year outlook. Our order book remains very strong and continues to exceed our ability to supply. Looking to next year, we expect solid growth based on the continued strong demand for our products and available foundry supply. We are excited about our trajectory this year. Applications that use SSD and eMMC+UFS are doing well. Adoption of these solutions is growing and, furthermore, we are growing our market share. Let me discuss key factors driving our growth this year. First, our SSD controllers. While the PC market is experiencing modest growth, the SSD attach rate continues to rise, which is beneficial for the growth of our SSD controllers. The primary reason our controllers are growing rapidly is due to market share gains. Several of our customers have been gaining market share, and we have also been gaining share among our customers. They have also been upsizing their orders to us as they consolidate their procurement. We have been gaining SSD controller market share and expect to continue this trend. We have sized the client SSD market and now believe our market share last year was in the 25% to 30% range. Based on our strong controller sales growth this year, primarily driven by PCIe Gen3 SSD and OEM programs, we believe we should pick up 5 to 10 percentage points of market share. Our market share is increasing because our PCIe Gen3 PC OEM programs, which we have discussed extensively in the past, are scaling. We will be shipping our Gen3 SSD controller to 6 of 7 NAND flash makers. We are also shipping our Gen3 controller to several module makers with PC OEM programs. SSDs using our Gen3 controller will be used by all of our global PC OEMs. Additionally, we are gaining market share because our small merchant competitors and the captive controller resources of NAND flash makers are both also facing foundry supply issues. Our momentum will pick up further when our PCIe Gen4 SSD controller begins to ramp. Our Gen4 controller will start shipping in the middle of this year, initially with a few PC OEMs using the Intel Tiger Lake platform. We expect our Gen4 OEM program to scale quickly next year when Intel's Alder Lake platform becomes widely available. Based on the size of our current Gen4 design win for OEM programs, when fully ramped, we believe our controller will be used in about half of our PC OEM Gen4 SSDs. Let me add that we continue to outgun the competition. Last year, our client SSD controller share in the merchant market was 4 times bigger than our closest competitor. And we expect our market share to grow further as more of our OEM SSD programs enter production and scale. Now turning to our eMMC+UFS controllers, which are primarily used in smartphone and IoT devices. For smartphones this year, we are benefiting from an industry recovery. Additionally, with smartphone embedded storage technology continuing to rapidly transition from legacy eMMC to UFS, our UFS controller sales will continue to scale rapidly as we further benefit from this trend. Furthermore, we expect our share gains to increase as our customers gain UFS market share. We believe our U.S. NAND flash partner using our highly competitive, jointly developed UFS controller with their industry-leading NAND and mobile DRAM technology will continue to outperform. We are also seeing positive customer engagement from our Chinese module makers and other customers with their UFS programs. In the legacy eMMC segment, we continue to be positively surprised by both the sustained power of eMMC and the expansion of our market opportunities. eMMC remains an ideal storage solution for applications that do not require large data storage capacity and are price-sensitive, and will continue to be widely used for the foreseeable future by low-cost smartphones as well as IoT and other non-smartphone applications, such as smart TVs, streaming devices, dongles, set-top boxes, smart speakers, and mainstream Chromebooks. Our eMMC controller opportunity has also grown significantly larger with NAND makers exiting the eMMC market as their priorities have changed. We are becoming the primary supplier of eMMC controllers. There are no other meaningful merchant suppliers of eMMC controllers. Our strong order book is a result of solid end markets and growing demand for our controllers. Demand for our controllers continues to exceed our ability to supply to our customers because of the very tight foundry industry capacity. We still do not have enough foundry capacity to manufacture all the products ordered by our customers. We have, however, secured additional foundry supply and as a result, are able to meet more of our customers' needs and increase our full year revenue guidance. We have also secured additional foundry supply that should enable us to deliver solid growth next year. But again, demand for our products will likely continue to exceed our ability to supply. We are thankful to TSMC and our customers for their support in securing additional foundry supply. And to our many customers with unfulfilled orders, we ask for their understanding in this matter. We continue to work on securing incremental foundry supply without which the many ecosystems, such as PCs and smartphones, that depend on our controllers for storage solutions could be negatively affected. Last quarter, we announced our $1 billion sales target by 2023 in 3 years. With stronger sales growth this year and solid growth next year, it is likely we could reach our goal sooner than our original expectation. Let me conclude by providing an update on our enterprise-grade PCIe Gen5 SSD controller. Engineering work continues, and we have been marketing our unique Gen5 architecture to hyperscalers in both the U.S. and China with very positive feedback. We expect to start sampling this SSD controller with customers in the second half of next year. Now I will turn the call over to Riyadh to discuss our financial results and our outlook.

Thank you, Wallace, and good morning, everyone. I will discuss additional details of our first quarter results and then provide our guidance. My comments today will focus primarily on our non-GAAP results unless otherwise specifically noted. A reconciliation of our GAAP to non-GAAP data is included with the earnings release issued yesterday. In the first quarter, revenue reached a record $182 million, 27% higher sequentially and 37% higher year-over-year. Earnings per ADS were $1.11, 29% higher sequentially and 38% higher year-over-year. Now I will walk through the performance of our three key products during the first quarter. SSD controller sales increased 25% to 30% sequentially and 45% to 50% year-over-year. Sequential strength in our first quarter SSD controller sales growth was primarily driven by our PCIe Gen3 OEM programs. Channel market sales also grew sequentially, but more modestly. eMMC+UFS controller sales also reached a record high, growing 30% to 35% sequentially and 50% to 55% year-over-year. Sequential strength in our eMMC+UFS controller sales growth was primarily driven by our UFS controllers. eMMC controller sales also grew sequentially, but more modestly. SSD solutions sales increased 0% to 5% sequentially and were down 40% to 45% year-over-year. Gross margin in the first quarter increased from 49.3% in the prior quarter to 50.7%. Gross margin was higher than our guidance, primarily due to our focus on selling a better product mix for non-OEM programs. Operating expenses in the first quarter were $43.9 million, $4.4 million higher than the prior quarter, primarily due to higher compensation accruals. Operating margin in the first quarter was 26.6%, an increase from the 21.9% we generated in last year's fourth quarter. Our effective tax rate in the first quarter was 20.9%, above our 15% to 20% tax rate guidance due to more sales from higher tax offering entities. Stock-based compensation in our operating expense, which we exclude from our non-GAAP results, was $3.0 million in the first quarter, slightly below our guidance of $3.1 million to $3.3 million. We had $371 million of cash, cash equivalents, restricted cash and short-term investments at the end of the first quarter compared to $369.2 million at the end of the fourth quarter last year. We paid $12.2 million in dividends to shareholders, the second quarterly installment of our $1.40 per ADS annual dividend that was announced last October. Now let me turn to our second quarter and full year guidance and forward-looking business trends. For the second quarter, we expect revenue to increase 5% to 10% sequentially to approximately $192 million to $201 million. We expect all three of our key products, SSD controllers, eMMC+UFS controllers, and SSD solutions to all grow in the second quarter. As Wallace discussed, our growth remains capped by foundry supply limitations. Second quarter gross margin should be in the range of 48% to 50%. Second quarter operating margin should be in the range of 26% to 28%. In the second quarter, we expect stock-based compensation in the range of $2 million to $3 million. For the full year 2021, we are now expecting the following: With additional foundry supply allocation, we now expect revenue to grow in the range of 45% to 55% to $782 million to $836 million. Our ability to meet customer demand, however, will remain capped by foundry supply availability. Our full year gross margin expectations remain in the range of 47% to 49%. We expect gross margin in the second half of the year to be less favorable as our back-end and substrate costs have increased, while our prices to customers for OEM programs are bound by contractual terms. Furthermore, our OEM programs are driving our growth, and despite foundry capacity limitations, scaling faster than expectations and subject to volume discounts. We believe our short-term investment in slightly lower gross margin will pay off in the years to come as we pull further ahead of both merchant and captive competitors and secure our position as the tech industry's primary supplier of NAND controllers. Operating margin is expected to be in the range of 26% to 28%, up sharply from 21.8% last year and approaching our 30% target. For the full year, we expect stock-based compensation in the range of $16 million to $18 million, more than the prior year. We expect our effective tax rate for the year to be about 20%, similar to the first quarter. On February 24, we broke ground on the construction of our Hsinchu office building. Construction is budgeted to cost $77 million, with $5 million spent this year and $33 million next year. We expect to complete construction in 2024. Upon completion, we plan on a sale and leaseback of the building. Separately, on February 18, we won a bid with a third party to build an office building in Taipei, paid a $1 million bid bond and expect to execute a property development agreement in the next few weeks at which time, we will pay a $5 million performance bond and other fees. We expect to spend approximately $75 million to $85 million and complete construction in about 6 years. This concludes our prepared remarks. We will now open the call to your questions.

Annie, please go ahead and give instructions for the Q&A session.

Our first question is from Anthony Stoss of Craig-Hallum.

Congrats on just the unbelievable results for you guys. A couple of questions. Riyadh or Wallace, maybe if you could give us your sense right now of when you think the wafer shortages really start to ease up? TSMC and others are talking about in Q3, things will start to loosen up a little bit. So I'm curious what your view is on that? And then also, either one of you two guys, you're talking about bookings being sharply above your ability to get wafers. Can you quantify that? Is it 30% higher, 50% higher? Or maybe just if you wouldn't mind sharing kind of your view on where your bookings are for this year.

Yes. I believe TSMC said they expect the current foundry shortage to continue throughout this year and maybe extend into 2022. As many of you know, the foundry supply shortage is especially acute with the mature technology nodes, which significantly affect our ability to have a product fabricated. None of our products today are manufactured using 5- or 7-nanometer technologies. We do not know when the foundry supply will improve.

Tony, let me add to Wallace for your other question about our bookings, your question about how much our bookings could be in excess of our wafer supply. One way to put a framework around that is if we do not have that constraint, we will be significantly on our way towards the $1 billion sales target that we outlined last quarter. We believe we'll be delivering faster than expected because of our stronger growth this year.

And then as a follow-up, your comments or Wallace's comments about solid growth for 2022. Do you think you could grow faster than the overall, let's say, PC market in 2022 over 2021 revenues?

I think for the demand, yes. But from the wafer supply, we are not sure. But we already secured additional wafers to support our growth, but the demand is much bigger than the supply we can secure today.

Congratulations on the real robust execution, guys. Great to see the growth and the margin performance has come through in the first quarter. So the first question is just on the new $1 billion target, and it's really a clarification. So the clarification is this: Can you just identify what the prior and what the new embedded underlying PC growth is in the calendar '21 revenue target, which at a midpoint, I think is $809 million? And is there any benefit from some of the improved product mix and pricing that you're seeing in channel markets in the increase from initial to $809 million now?

Okay. Let me start with your first question. The $1 billion target, the underlying assumptions that we had baked in was our growth will be driven primarily from our SSD controllers and also from our eMMC+UFS controllers. The growth from these two products will get us to $1 billion. Additionally, we also expect our FerriSSDs to contribute, and we are not expecting our enterprise SSDs to be part of this. Our enterprise SSD, we're building as a growth driver beyond the $1 billion. In terms of PC growth, we are obviously expecting modest PC growth this year, in line with expectations from third-party industry analysts. And within the PC profile, we are expecting notebooks to grow, while desktops to decline. And so that is a favorable tailwind to our overall business. But clearly, as Wallace had talked about, the key driver is not the PC growth or the transition from hard disk drives to SSDs. But rather because of our investments that we had put in place over the last few years in winning design wins with our big customers for OEM programs, these are now going into production and scaling very nicely, leading to very solid market share gains.

Yes. Good to see those gains. And Riyadh, a follow-up just on the full year '21 guide implications for some things in the back half. So on revenue, just given your updates into increased supply and the share gain programs you have in Gen4, would you expect revenues to rise sequentially through the back half of the year? Or do you get supply early enough that we would see revenues peak in the third quarter? And then on the gross margin side, the full year gross margin guidance on average would imply 46.5% gross margin in the back half of the year. But as you've noted, gross margin is very mix dependent. And so what does the order book portend for the way gross margin can play out in the back half?

Okay. For the first part of your question, our revenue expectations are for very modest sequential revenue growth for the rest of the year, quarter after quarter, very modest growth. This is going to be in line with additional wafers made available to us. But clearly, our ability to grow is capped by wafer limitations from our foundry suppliers. In terms of our gross margin, our gross margin profile is going to be less favorable in the second half of the year as our large OEM programs continue to scale. And as these OEM programs continue to scale, our gross margin will be affected. For many of our OEM programs, the pricing has already been predetermined and baked in contractually, so our flexibility in terms of changing prices is fairly limited. Furthermore, there also volume incentive plans. And as these programs get larger and larger, they will impact our gross margin negatively in the second half of the year.

So let me comment. I think all wafers are available right now. We won't gain any additional wafers. That's why we make a full year guidance based on all the available wafers to us. Now execution about manufacturing back end, all the assembly, testing, everything, and we know exactly the yield, and the price is all fixed. That's why I think the guidance, we'll fulfill the full year. It just depends on the rest of our operational execution.

Okay. That's fair, guys. Would it be fair to think that your OEM customer mix would rise from the third to the fourth quarter and, therefore, we would see a sort of gross margin impact to the business sequentially?

That is correct.

Congratulations on the excellent results. Wallace, you mentioned that your market share is improving significantly, with an expectation to gain 5 to 10 points. Could you elaborate on what's driving these market share gains? How sustainable do you think this is heading into next year? Additionally, is this a major factor in the underlying growth in SSDs and your confidence in reaching the $1 billion target?

Okay. It's a very good question. I think we have very high confidence to gain 5% to 10% market gain because of our current PC OEM program from NAND makers as well as module makers. And we are ramping very quickly with the available wafers to us. And due to, as we mentioned, our PCIe Gen4 controller design win being very widely aligned with Intel Alder Lake next year. So we believe when PCIe Gen4 SSDs are adopted widely by PC OEMs, we will be about 50% of our OEM project. That's why we have very high confidence that we will have continued market gain for our clients continuing into next year and even beyond.

You mentioned that based on this year's sales growth and expectations for next year, you may reach the $1 billion target sooner rather than later. You indicated that fiscal year '23 could see this target, potentially as early as 2022. I'm trying to understand this in light of your comments about ongoing supply constraints at TSMC for the rest of this year and possibly into 2022. Are you anticipating additional wafer supply in 2022 that would boost your confidence in reaching the $1 billion target? Or is it that the Gen4 programs are set to ramp up, and having those purchase orders secured is providing some assurance that you could hit that figure sooner? I'm interested in how you reconcile the potential for quicker achievement of that target with the still significant capacity constraints.

So maybe we can make it very clear to you guys. Our backlog that's building today is over $1 billion. If we have sufficient wafer supply, this year, we can achieve over the $1 billion sales target. However, we are unable to do so due to wafer supply shortages. Now we have secured additional wafer supply, which is expected to support our growth for 2022. It's too early to comment on the 2022 guidance, but we believe we'll continue to grow for 2022. If we can continue negotiating with TSMC to support major critical programs, because if we cannot get sufficient supply, that will impact the ecosystem negatively. So this is very important for us and also for our customers and for the whole industry. So we believe the current secured additional wafers will support our growth for 2022, but we cannot comment on what the eventual sales revenue will be. However, we believe it will likely be earlier than 2023.

That's very helpful. For my follow-up, I'll step back in the queue. Riyadh, the gross margin is dipping to around 46.5% in the second half. You mentioned that pricing is fixed or predetermined, but we're also seeing an increase in back end and substrate costs. This situation seems temporary. I'm curious about when you expect those substrate costs to start decreasing. Will that still depend on the capacity-constrained environment? How are you balancing the pricing commitments for 2022 with the increasing back end and substrate costs as we look beyond the second half?

That's a very good question, Raji. This is also tied in with, when do we think foundry supply will ease up. So related to this topic is when do we expect substrate and back-end tightness to also ease up. Frankly, it's not that clear right now. We are seeing improvements on the back end and substrates. But when this will lead to a better cost structure in terms of pricing of back end and substrates to us, we're still not quite sure. We are keeping a very close eye on this and continue to see if we can negotiate better terms. This is something we'll be updating everyone in the following quarters.

Two questions, please. I guess, Riyadh, how sustainable are these OEM programs? You've indicated contractual agreements with NAND OEMs, but I'm curious if you could discuss the volume commitments you have with the OEMs that extend into the second half and into 2022 in the event capacity constraints ease. And how that may increase the risk that those OEMs move back toward internal controller solutions?

Yes. I think these OEM programs are not easy to be replaced by either a third-party controller or internal controller. Because it needs to be pre-designed and must undergo significant effort in the design qualification and customer verification. I think we already have a full year forecast from 2022. So we know exactly what is going to last. This is not just a solution. This is an industry standard and will continue to stay for more than 3 or 4 years. I think we have a contract price that has been discussed 2 years ago. Nobody could predict the COVID-19 and the wafer shortage and all these manufacturing-related issues. But we respect the commitment regarding our customers. We have been working together for a long time, and we committed to further agreements and contracts. So we think this program is going to empower overall gross margin. However, this is going to drive further growth and net profit for the company. This will be a unique, outstanding high-growth project for the industry. So we're trying to adhere to our previous contracts and agreements and moving forward.

Karl, let me also add a little bit more to what Wallace said. As Wallace mentioned, these programs took a lot of effort and a long time to execute. Our Gen3 projects, our PCIe Gen3 projects, we've been working on these for over 2 years, and they are now finally going into production and beginning to scale. As a result of the work that we've been doing with our customers for these OEM programs, we're building a lot of traction, a lot of stickiness. This is now leading to the Gen4 programs that we have. This gives you a sense of the solidifying position as a leading supplier of controllers to the tech industry.

Yes. No, that's very helpful. For my follow-up, how would you characterize the level of channel inventory of controllers today? I ask because I think there have been perhaps some news flow that module makers have been stockpiling PCIe controllers as an afterthought or as a view that cloud SSD demand remains strong given this recent uptick in cryptocurrency demand. So any thoughts there in terms of the level of channel inventory of controllers today and your ability to service that demand?

So let me comment. This is a good question. I believe that the current channel controller inventory is very low, probably almost near zero because everybody is in shortage today. The NAND component supply is not in shortage. So no customer is trying to buy additional inventory or solutions to stock. I think every product, every wafer, every component we make is shipped directly to the end customer. Our agents and distributors cannot even keep it in stock. We see inventory is very, very low. Demand depends on the product. Some products have really high demand, while some have moderate demand. But in average, I believe there's almost no inventory or the inventory is very low for controllers in the channel.

And congrats on the good results. The first question is also related to gross margin. So for 2022, do you think there is a chance that we can renegotiate the price with the OEM customers in terms of revising up the price, as you have said that the foundry supply could be pretty tight all the way into 2022? And also, could you clarify the so-called OEM project includes both SSD controller as well as UFS controller or mainly SSD controller?

I think that everything is negotiable, but for a very large program with a contract, it's difficult and challenging. Especially, I think for some additional wafers, we could also work with our major OEM customers to negotiate together. We respect the contract and engage with the customer. But I think for 2022, we will definitely try our best for the company. We'll also see what the price trend for the wafers is. If wafer prices increase dramatically, we will have to negotiate the price with the customer. By the way, this OEM program involves both PCIe and UFS.

Donnie, let me also add, there are two other variables that we also look at. The first is, we're constantly winning new designs, right? With new designs, new contracts, new OEM programs, the terms could be more favorable, more reflective of current circumstances. The other factor is that while back end and substrates are very tight right now, the industry is always cyclical. Our semiconductor industry is always cyclical. So sometime down the road, there will be more supply, more back end, more substrates, and prices will change. Our gross margin circumstances will also become more favorable. These are two other factors to keep in mind.

Thank you, Riyadh. I appreciate your input. I have a follow-up question. When you talked about back-end capacity and substrate supply, how does that compare to the foundry capacity constraints? Which one do you expect will see an easing of supply tightness first? Additionally, I understand that Samsung’s Austin fab faced challenges due to severe weather but has since resumed some production for controllers. Have you observed any changes in supply and demand since they restarted their production?

In my personal opinion, the back end, the substrate supply and the assembly capacity, that will be relieved earlier than the wafer supply. I think when we talk about wafer shortages, we need to separate advanced technology nodes from mature technology nodes. Currently, I believe, 90% of the companies are using mature technology nodes, and increasing capacity for these wafers is quite hard. It takes a minimum of 2 years to land, build a clean room fab, and install new equipment to fine-tune. It will take time. So it won't result in an increase in capacity within just 1 to 2 years. The tight wafer supply situation will continue, but it also depends on whether the relief is related to overbooking or other factors, which might resolve certain supply tight spots. However, according to TSMC, they don't foresee a solution this year, and probably the tight capacity will extend into 2022.

That's very helpful. And the last question is regarding the operating margin. So it looks like the gross margin, of course, in the second half, there could be some pressure. But for operating margin, it looks like pretty strong. So Riyadh, just wondering, besides operating efficiency, are there any other reasons behind the improving operating margin?

Donnie, it's a big trade-off that we have, right? With these OEM programs, they're very large. And when they ramp, the trade-off is that we have these volume incentive programs that result in slightly lower gross margin, but the offset is we're getting very significant operating leverage, which helps our bottom line significantly.

Yes, this is actually Mehdi Hosseini. I have a couple of follow-ups. And by the way, congrats for gaining market share. Wallace, how are you seeing new cryptocurrency technologies, specifically Chia? There have been a lot of headlines over the past month as to how the new cryptocurrency is going to consume low-end SSDs. It appears that perhaps some of the channel inventory of low-end SSDs has been completely depleted. I'm just wondering if you have any thoughts you can share with us? And I have a follow-up.

Yes, I think the new crypto technology and Chia, the preference really is for capacity drives. It's really easier for HDD, our high-density enterprise SSD, not for client SSD. So this is really not related to our business. We didn't pay much attention. I think even in our channel, we're focusing on the real customers to grow our business and technology. But we know this relates to the high-density HDD and enterprise SSDs.

Mehdi, apologies, I appreciate if you could repeat your second question.

Yes. I was going to get into it. Actually, two-part second question. One for Wallace. You have historically had pretty good assessment of NAND supply and demand. You just told us that there is absolutely no channel inventory of controllers. Does that apply to NAND, especially looking into the second half? And the second part of the second question is for Riyadh. Would it be fair to assume that perhaps OpEx is going to be in the range of $40 million to $42 million on a quarterly basis for the foreseeable future?

I think let me comment regarding the channel for controller and the NAND. As I said, we believe there is no inventory for controllers because every controller we have in hand will be sold out. We have a huge backlog in hand and cannot fulfill the orders. Regarding the NAND, NAND makers also observe there's a severe controller shortage. So NAND makers moved their storage products to higher density products, which will consume more NAND. As you can see from PC OEMs before, the lower density, 128 gigabyte SSDs are not in demand anymore. So the value line starts from 256 gigabytes. That's how NAND makers manage to move all the products to higher density, and that can consume NAND quickly. That's why we don't see that much inventory even in the NAND market. From Q2, we start to see NAND prices going up. I think the price will continue to gradually increase but within single digits. So we really don't see high inventory, either for controllers or storage solutions in the channel.

Mehdi, to your other question about our OpEx in the second half of the year. We are expecting our OpEx to be fairly stable sequentially in the following quarters. We're expecting our top line to inch up a little bit quarter after quarter. We're expecting our gross margins to drift down a little bit, but we are expecting our operating margins to be fairly stable. So this is going to be coming from fairly stable operating expenses in each one of the quarters in the second half.

Congratulations on the results here. As you approach that $1 billion revenue, can you give us a sense of what kind of mix of SSD and smartphone eMMC+UFS controllers we should expect roughly?

We cannot really tell the detail of percentage. We just know that both the client SSD and eMMC+UFS are very strong. We also see our Ferri product going strong in the second half of this year.

Okay. That's helpful. And then for UFS, can you talk about the number of NAND customers that are now ramping? Is it still maybe 2 major ones? Or is it more at this point? And how many do you expect 6, 12 months out?

So for this year, the momentum is really driven by one major U.S. NAND maker. But we're going to have 5 more customers to grow UFS products in 2022.

So all five of those will be starting, Wallace, in 2022? Or some come on this year?

For the year, it will be very small, very light looking. But all five will really ramp in 2022.

As I think Wallace talked about shipping to 6 of the 7 NAND fabs. Can you remind us where that was a year ago? Also, can you characterize which markets you're getting more traction with your Gen3, Gen4 solutions? Is it primarily product ending up in retail? Is it low-end OEM? Is it across the gamut? And I have one follow-up.

As we mentioned earlier, we currently hold about 70% market share among module makers. However, last year, our OEM program was slightly below 30%. This year, we plan to expand our PC OEM program. The main driver will be the PCIe Gen3 controller, which is seeing strong demand from PC OEMs and is helping us gain market share. We anticipate this momentum will carry over into Gen4 next year, as we have approximately 50% design win sockets in our PC OEM Gen4 program that align with the Intel Alder Lake platform for 2022. We expect to maintain this momentum and continue to increase our market share, leading to a rise in our PC OEM percentage from 2021 into 2022.

Matt, the line was a bit choppy. I appreciate if you could repeat your first question again.

I think, Riyadh, that Wallace got most of what I was asking for. Sounds like it's predominantly PC OEMs. But my follow-up is, is there any color you can give us around how much the newest NAND makers' growth is contributing to your strong results and better outlook moving forward?

The new NAND makers' OEM program will contribute from 2022.

Guys, I wanted to go back to the eMMC market. And it's typically not a focus, but there were 2 notable things in the commentary. I wanted to follow up on one. There is an instance where it appears one OEM, due to some of the recent dynamics that we're seeing in the marketplace and with their manufacturing, may be moving away from doing eMMC internally. And you mentioned that you're the only merchant eMMC supplier. So the twofold question is this: one, is it possible that the decision around eMMC could be the tip of the iceberg for more controller work, which would actually go to merchant, not just eMMC, but SSD controllers? And two, given the industry structure with merchant eMMC now, how would you characterize the pricing dynamics? Is it more possible to price for functional value and the value you create, given the industry landscape?

Okay. Let me try to answer your three questions. First, regarding eMMC, it's a very good question. I'd like to answer it. I think the market trend favors Silicon Motion in the next few years. Firstly, the market demands more low-density storage solutions. eMMC is the best solution for lower density storage. Lower density means 120 gigabytes or lower. Because of the controller shortage, the wafer shortage has become a controller shortage; NAND makers have tried to utilize their variable controllers to make higher-density solutions. That's why NAND makers are moving away from eMMC controllers to UFS or higher-density SSDs or other storage solutions. This gives us tremendous opportunity. We see tremendous demand coming to our company, but we cannot even fulfill 50% of the demand because it is so large. Even more NAND makers are coming to us to place purchase orders, and we have to apologize because we do not have the wafers to support the trend. We expect to see a huge demand for eMMC, as there are no other significant merchant suppliers with the capability and scale to support them. Second, we see this will happen and continue for mainstream client SSDs. Because NAND makers finally realize they don't have to develop the client SSD controller for mainstream or value lines. They will focus on high-end client SSDs or enterprise SSDs for their own benefits. Our turnkey solution is more cost-effective, time-to-market, and serves our customers with a very integrated technology. This is better than internal development. Regarding the third question about pricing, when wafers are in short supply, prices will stabilize, which favors our controller side. We will, based on the product, make certain adjustments. If a certain product was below our corporate average in the past, we will do a little adjustment. But if this is contract-based for a major OEM, we will respect the contract.

Furthermore, as Wallace talked about, more and more new businesses are coming in our door. To the extent that we can take on new projects, the terms of new projects will be reset at today's conditions. The margins for new projects will be more favorable than projects that we took on, say, 2 years ago, facing higher costs today.

And as there are no further questions on queue, I'd like to hand the conference back to our President and CEO, Mr. Wallace Kou. Go ahead, please.

Thank you, everyone, for joining us today and for your continuing interest in Silicon Motion. I would leave you with some final thoughts. I'm extremely proud of our execution over the past several quarters and want to thank our team for their hard work and dedication. As a large merchant controller supplier, we play an increasingly critical role in both the NAND supply chain and many critical technology ecosystems. Technology today is built on data, and data needs to be stored to be used. Solid-state storage requires controllers; our competitive position in the storage controller market has never been stronger, and I remain confident that our growth will continue. Thank you for your continued interest and listening to our call. Goodbye for now.

Thank you. And this concludes today's conference call. Thank you for participating. You may now disconnect.