慧榮科技 (SIMO) 2018 Q2 法說會逐字稿

This conference call contains forward-looking statements within the meaning of Section 27A of the Securities Act of 1933 and Section 21E of the Securities Exchange Act of 1934 as amended.

Such forward-looking statements include, without limitation, statements regarding trends in the semiconductor industry and our future results of operations, financial condition and business prospects.

Although such statements are based on our own information and information from other sources we believe to be reliable, you should not place undue reliance on them.

These statements involve risks and uncertainties, and actual market trends and our results may differ materially from those expressed or implied in these forward-looking statements for a variety of reasons.

Potential risks and uncertainties include, but are not limited to: continued competitive pressure in the semiconductor industry and the effect of such pressure on prices; unpredictable changes in technology and consumer demand for multimedia consumer electronics; the state of any change in our relationship with our major customers; and changes in political, economic, legal and social conditions in Taiwan.

For additional discussion of these risks and uncertainties and other factors, please see the documents we file from time to time with the Securities and Exchange Commission.

We assume no obligation to update any forward-looking statements, which apply only as of the date of this conference call.

Ladies and gentlemen, thank you for standing by, and welcome to Silicon Motion Technology Corporation Q2 2018 Earnings Conference Call.

(Operator Instructions) I must advise you that this conference is being recorded today, Wednesday, 1st of August, 2018.

I would like to hand the conference over to your first speaker today, Mr. Jason Tsai.

Thank you.

Please go ahead.

Thank you, and good morning.

Welcome, everyone to Silicon Motion's Second Quarter 2018 Financial Results Conference Call and Webcast.

My name is Jason Tsai, and with me here is Wallace Kou, our President and CEO; and Riyadh Lai, our Chief Financial Officer.

The agenda for today is as follows: Wallace will start with a review of our key business developments.

Riyadh will then discuss our second quarter financial results and provide our outlook.

We'll then conclude with Q&A.

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 involved in investing in our securities, please refer to our filings with the U.S. SEC.

For more details on our financial results, please refer to our press release, which was filed on Form 6-K after the close of market yesterday.

This webcast will be available for replay on our website, www.siliconmotion.com, 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 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.

With that, I will turn the call over to Wallace.

Thank you, Jason.

Hello, everyone, and thank you for joining us today.

I will first update you on our business, and then Riyadh will review our financials and provide our outlook later on this call.

We are pleased to report second quarter revenue of $138.1 million, which is up 6% sequentially and up 4% year-over-year.

Earning per ADS this quarter was $0.92, an increase of 29% both sequentially and year-over-year.

This quarter, our SSD controller grew approximately 5% sequentially.

With rapid growing NAND supply and increasingly variable NAND prices, flash maker are more focused on SSD opportunity and raising their sale projections.

As a result of flash partners' improved SSD growth outlook, we are increasing our full year SSD controller sales growth to at least 30%, with sale accelerating in Q3.

Additionally, during this quarter, we grew our pipeline of SSD controller project at our 3 NAND (sic) [3D NAND] flash partners, Intel, Micron and WD, by over 50% sequentially and more than double compared to last year.

The majority of our pipeline of new SSD program are for PCIe, NVMe and now include those are used 64, 72 and 96-layer TLC and QLC 3D NAND with production running from this quarter all the way through 2020.

NAND prices have fallen meaningfully since the start of this year.

Dollar per gigabyte cost of NAND has been falling and will continue to fall rapidly with current high-volume 64 and 72-layer 3D NAND production.

Scaling of 4 bits per cell QLC NAND and migration to 96 layer and upcoming higher layer count NAND.

Since the NAND industry is very competitive with now 7 suppliers, the benefit of increasing lower NAND costs is quickly passed to device OEM and consumers.

We continue to believe that as NAND costs fall further, SSD demand will accelerate further.

We have been broadening our portfolio of SSD controllers and deepening our relationship with NAND flash makers with our highly configurable hardware plus firmware platforms and with module makers with our turnkey solutions.

These solutions range from low cost to mainstream to high-performance market segment DRAM-less and with DRAM, SATA and PCIe NVMe.

In last year Flash Memory Summit, we introduced a series of new PCIe NVMe client SSD controller, high-performance SM2262, mainstream SM2263 and DRAM-less SM2263XT.

We are pleased that these controllers are all in high-volume production and SSD using our controller benchmark is ahead of the class, including, again, Samsung SSD and have been receiving favorable third-party reviews.

Example of these controllers include HP's flagship EX920 SSD using SM2262, with Intel 64-layer 3D flash and HP's entry-level DRAM-less NVMe SSD EX900, using SM2263 XT with Micron 64-layer 3D flash.

Our SM2262 and SM2263 family of controller are key driver of our growth in the second half of this year.

This August, next week, at the year's Flash Memory Summit, we look forward to introducing our next-generation SSD controllers.

Let me now update you on our high-performance SSD controller designed specifically for data center applications.

These controllers support both open-channel as well as standard NVMe implementations and will first rollout at 2 leading Chinese hyperscalers for their open-channel requirements.

As previously discussed, we are in the late-stage, preproduction testing at these 2 hyperscalers.

And separately, we are now a technology partner of Microsoft's Project Denali, which is having open-channel SSD application for the open compute projects.

We expect both Chinese hyperscalers to complete testing of our open-channel SSD in the third quarter.

And as testing is completed successfully, production to begin immediately afterward.

And this will be world's first open-channel SSD controller in commercial production.

We expect that when production begins, initial sale will certainly be fairly small, but demand will grow meaningfully next year.

Increasingly, the global leading hyperscalers, given their considerable scale and unique technology requirements, are taking greater control of their SSD requirements.

They are independently procuring NAND flash components and SSD controllers, designing their own SSD and outsourcing to contract manufacturers.

We believe we are very well placed as a supplier of high-performance, highly configurable hardware plus firmware, data center-class SSD controller solutions to hyperscalers.

We can directly supply controller solutions, whether for open-channel SSD or standard NVMe SSD.

Or as in the case of Alibaba and other Chinese hyperscalers, SSD controller solutions plus SSD solution design and driver software development.

Now let me turn to our eMC (sic) [eMMC] and UFS controllers.

Sales declined slightly in the quarter, we believe, in line with our flash partners, all eMMC sales patterns in this quarter.

China smartphone sales were weak, and we believe our flash partner was more focused on SSD.

Last quarter, we mentioned that embedded memory of Android smartphone have begun migrating from eMMC to UFS.

We are seeing increasing smartphone design activity using UFS, especially at a -- premium flagship devices.

We believe that later this year, Qualcomm and MediaTek will likely launch new upgraded processors for mainstream smartphones that support UMCP, single-chip embedded memory that contain NAND flash, mobile DRAM and the UFS controllers.

We are encouraged by our growing sale of UFS controller to Micron in the second quarter.

And we believe UFS controller sales could account for a meaningful proportion of our overall eMMC plus UFS controller sales in the second half of this year and continue to grow strongly in 2019.

Moving to our SSD solutions.

Second quarter SSD solution sales increased sharply as our new project in Alibaba scales.

In the third quarter, however, our Shannon Data Center SSD will be [considerably] weak.

The third quarter is typically a strong quarter, but this will not be the case this year.

As you know, sales of our Shannon SSD tend to be lumpy.

In the third quarter, large SSD sale to our primary hyperscale customer have been delayed to the fourth quarter due to new procurement timing.

We expect fourth quarter Shannon SSD sale to rebound.

Our other SSD solutions, our Ferri industrial SSD, also grew in second quarter and expected to grow further in the third quarter.

Our Ferri product address a diversified set of application for all this automation to commercial equipment and to server (inaudible) storage for OEM Japan, the U.S. and China.

Recently, we have added solid growth from sales to automotive component manufacturers that are using our automotive-grade FerriSSD in infotainment systems for German and Japanese car brands.

Let me conclude by saying that we believe we are now beginning to see demand for client SSD increasingly and meaningfully from price [erratically] of demand.

NAND prices have fallen meaningfully and SSD demand.

And as it relates specifically to us, demand for our SSD controller have increased.

Our flash partner also benefiting from this trend and so have therefore sharply expanded their SSD controller project with us.

PC, game console, other client devices and external storage still account for more than 300 million client hard disk drive this year.

And as NAND prices continue falling from the introduction of increasingly low-cost NAND component over the next few years, we strongly believe that most, if not all, of this HDD will be replaced by SSDs.

I will now turn the call over to Riyadh to discuss our financial performance and outlook.

Thank you, Wallace, and hello, everyone.

I will summarize our financial results and provide our outlook.

Before I begin, I would like to reiterate that our 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.

Our Q2 revenue increased 6% sequentially, in line with guidance.

For Q3 2018, we expect revenue to be in the range of $136 million to $143 million, representing a decrease of 1.5% to an increase of 3.5% sequentially.

For full year 2018, our outlook remains unchanged, and we continue to expect revenue to increase 5% to 10%.

We're expecting stronger-than-expected SSD controller growth in the second half of this year, but this strength will be offset by 1 quarter delay of large Shannon sales with our leading hyper-scale customer.

I will now walk you through the performance of our key products before taking you through key elements of our P&L and the rest of our financials.

In Q2, sales of our SSD controllers grew approximately 5% sequentially.

In Q3, we expect our SSD controller sales growth to accelerate meaningfully as new PCIe and NVMe programs begin to ramp.

SSD controller sales should grow further in Q4 and for the full year 2018.

We believe our SSD controller sales will likely grow at least 30%, meaningfully higher than the 20% growth rate that we had previously communicated.

In Q2, sales of our eMMC plus UFS controllers declined approximately 5% sequentially.

In Q3 and for full year 2018, we expect our eMMC plus UFS controller sales to be in line with 2018 smartphone industry forecasts, which is likely flattish.

In Q2, sales of our SSD solutions increased approximately 35% sequentially as Shannon SSD programs at Alibaba ramped and diversified FerriSSD sales continue to grow.

In Q3, we expect our SSD solution sales to decline sharply as large orders are delayed 1 quarter and rebound in the fourth quarter.

As a result of this delay, we no longer expect our SSD solutions to grow at least 20% for the full year.

Instead, our SSD solutions will at most grow 20% for the full year.

Sales of our memory card and USB flash drive controllers and other legacy products will decline as previously communicated for full year 2018.

Moving to gross margins.

Gross margins decreased sequentially from 48% in Q1 to 47.5% in Q2 due to strong sales growth of low gross margin SSD solutions relative to other products.

This trend will reverse in Q3 when our low-margin SSD solutions sales decline sharply and growth of our high-margin SSD controllers accelerate.

Q3 gross margin is expected to be in the range of 48.5% to 50.5%.

Our full year gross margin will also benefit from stronger-than-expected, high-margin SSD controller sales and less lower-margin SSD solutions sales.

We now expect full year gross margin to be slightly higher than previously guided and in the range of 47.5% to 49.5%.

Our operating margin increased from 22.1% in Q1 to 25.7% in Q2 because of lower-than-planned operating expenses, primarily from lower R&D spending.

For Q3, we expect operating margin in the range of 25% to 27% as higher gross margin -- higher gross profits are offset by higher operating expenses for more R&D project tape-outs.

For the full year, we are increasing our operating margin guidance to the range of 24% to 26%.

Total headcount at the end of Q2 increased to 1,273, which was 11 more than at the end of Q1.

Our effective tax rate in Q2 was 10%, lower than the 14.6% in the previous quarter.

We expect our effective tax rate for the full year 2018 to be approximately 15%.

Our Q2 EPS was $0.92, significantly higher than $0.71 last quarter and $0.71 a year ago.

In Q2, stock-based compensation in our operating expense, which we exclude from our non-GAAP results, was $0.7 million, lower than the $3.1 million in the previous quarter due to the seasonal timing of RSU awards.

For Q3, we expect stock-based compensation to increase to between $5.4 million and $6.4 million and for full year 2018 stock-based compensation between $16.6 million and $18.6 million.

We had $364 million of cash, cash equivalent and short-term investments at the end of the second quarter, $18 million more than in the previous quarter and $58 million more than a year ago.

In Q2, we paid $10.8 million of dividends to shareholders, the third $0.30 per ADS quarterly installment of our annual $1.20 per ADS dividend that was announced in October of last year.

In Q2, we did not repurchase any shares related to our $200 million share repurchase program.

This 12-month share repurchase program was authorized in July 2017 and expired this month.

Let me add that while we have repurchased our shares in the past, our primary means of returning capital to shareholders is from our long-established dividend program.

This concludes our prepared remarks.

We will now open the call to your questions.

(Operator Instructions) Our first question comes from the line of Karl Ackerman from Cowen.

Two questions, if I may.

First is certainly on the eMMC business.

So I know the current view that eMMC should remain the workhorse for mobile NAND across many of the mid- and high-tier smartphones.

However, you mentioned briefly on UFS adoption.

And I guess, I wanted to push a little bit more on that.

I mean, how should we think about that now in the back half of 2018 and also into 2019 and then frankly, therefore, your opportunity to diversify your revenue beyond your #1 customer, given your leverage toward -- and given, I guess, leverage toward Android smartphones in Asia are now using higher capacities of DRAM and to a lesser extent NAND to differentiate their devices to the consumer.

So regarding your question, it really depend on when UMCP become the mainstream product for smartphone.

We believe next year both Qualcomm and MediaTek to mainstream application processor and really will support UMCP.

And we believe UMCP will become popular in the second half of 2019.

And when the leading smartphone makers start to adopt the UMCP, then we believe that will favor our UFS controller shift into major NAND customer.

Understood.

I appreciate that.

I guess, as my follow-up question, so in your prepared comments -- well, they clearly underscore your enthusiasm for your client SSD and custom SSD opportunity even through December this year.

I guess, could you quantify how much of your implicit guide for December is already accounted for with your bookings today, particularly in your SSD solution business?

Could you repeat your question again, just to make sure that we understand what you're asking?

Yes.

So it sounds like the Shannon business is taking a dip in Q3, but you expect that to rebound in Q4.

I guess my question is, how much of it -- that business pick up in Q4 is in bookings today versus perhaps your bottom-up view of that rebound in Q4?

How we think about the visibility into that?

Well, in Q3 of this year, we were originally expecting growth rate -- top line growth rate of at least 10%, but that obviously is not happening, given that our Shannon sales have shifted to Q4.

But at the same time, we are expecting very strong growth of our SSD controllers in Q3, so there is some offsetting relating to those 2 parts.

In addition, we strongly believe Shannon SSD solution will pick up on Q4 so the program due to this 1 quarter delay.

Our next question comes from the line of Mike Burton from Benchmark.

So nice to see raising the guide for client SSD.

A 2-parter here.

Do you see this more as a share shift or the tide kind of raising all boats?

And what is your view of market penetration at this point?

And then secondly, have you begun to see any meaningful pickup in higher density client rise due to the falling NAND prices, or is this more of a return to normalcy in the supply and demand?

We believe the reason we really have a more design pipeline from NAND maker is due to the NAND price decline sharply, and NAND supply become more sufficient to the market.

So I think [we'll dual mode].

One is helping NAND maker with their SSD project, also help moving the potential extra inventory for the NAND.

Second is design pipeline increase due to the PCIe transition becoming strong in PC OEM.

And in Shannon today, I think the PCIe is just small portion, but in PC OEM, new design majority will be PCIe solution.

So our PCIe controller take (inaudible) to the market become very mature, and we support all type of 3D NAND from all NAND makers.

And second, we believe the pipeline -- the design pipeline in production will continue through 2019, even some go to 2020.

So our visibility of the client SSD controller sale is very clear and very strong.

So we are very confident to guide our SSD will increase [50% to 30%] for this year.

Our market share, we believe, will be over 30% for 2018.

Great.

And then, can you talk -- any movement on ASPs in mobile or client as [we see] NAND prices falling?

And then, I know that, typically, that's not the case, but then also any impact from greater PCIe adoption or QLC going forward on ASPs?

Our ASPs for client SSD controllers remain quite stable.

Obviously, we have ASPs declining for other legacy parts like our SATA SSD controllers, but we also have the new PCIe, NVMe controllers blending in at a higher ASP.

But overall, the blended ASP has been very stable.

Our next question comes from the line of Sujeeva Desilva from ROGH (sic) [Roth] Capital.

It's from Roth Capital.

Riyadh and Wallace, so the SSD solution delay from 3Q to 4Q, can you help us size on the magnitude of that?

And whether this is a new pattern going forward, 2Q and 4Q being the peak versus 2Q, 3Q, or it's just a onetime event?

This is a onetime event.

Originally, as I mentioned, we were expecting our Q3 revenue to grow over 10% sequentially.

But because of our Shannon sales delay, sequential growth will now be essentially flat.

With this large Shannon sales now taking place in Q4 -- so contrary to what we typically would expect of a weak Q4 for Shannon based on past seasonality patterns, we're now expecting strong Shannon sales growth in Q4.

It's too early to say whether this will be the new pattern going forward, but this is obviously a very different pattern for 2018.

Okay.

That's helpful.

We'll keep an eye on that.

And then the open-channel.

I think you talked about 2 China hyperscaler wins.

Are those customers new to SIMO, or have you already been shipping product to those guys?

We will be shipping preproduction volume for late-stage testing.

Yes.

One of our customer is Alibaba, the other is another Chinese BAT hyperscaler.

Okay.

Great.

And what's the size of that open-channel opportunity over time versus the SSD solutions business you have now, just to compare the 2 long term?

Open-channel, I think they will become very strong demand in China market.

And China (inaudible) player, they're really moving to more open-channel, especially from BAT and some private carmakers.

We think that will bring significant sale revenue growth for long term in China.

In U.S., we believe Microsoft is a leading (inaudible) open-channel.

They try to make it a standard for data center SSD.

I think all other major player, they all have a so-called open-channel like or their own standard, [but at --] this model is changing.

And as an independent controller maker with strong knowledge in the NAND as a turnkey firmware solution, we can tailor other solution for deepened demand diversification, give us tremendous opportunity to grow even in the U.S. market.

Our next question comes from the line of Rajvindra Gill from Needham & Company.

I echo the congrats on the good SSD growth rate.

So Wallace, you mentioned that NAND pricing is coming down meaningfully, and you're seeing attach rates accelerate for client SSDs.

I'm wondering if you can maybe characterize how much the pricing has dropped on a per gigabit basis and versus, say, 3 quarters ago?

And also, if you could give a sense of the attach rates for client SSDs?

How they are rebounding from last year?

Any sense there in terms of number?

Yes.

We believe NAND price have already fallen meaningfully since start of this year.

And we expect the trend to continue through the rest of the year and through 2019.

Samsung had publicly announced they predict about 20% to 30% price decline.

We cannot comment the percentage, but we believe this -- in the first half of this year, the NAND price decline modestly, and second half will continue to decline.

And I think flash maker, they continue to bring increasingly lower-cost NAND to the market, so dollar per gigabyte cost fall rapidly, and it will trigger the stronger demand for consumer side, also for the PC makers.

And also that will -- for NAND maker to rethink how to speed up to move their asset inventory because yield will become more mature for 3D NAND.

And (inaudible) become stronger.

And that's why SMI, our turnkey solution, will play a very important role.

That's why we see our design pipeline not just from NAND maker, also module maker is increasingly quarter-by-quarter.

That give us opportunity to have a strong outlook, even stability about our client SSD controller sales in the next few quarters.

Let me also add.

I'm sorry -- let me just also add that SSD pricing continues to improve as NAND flash prices continue to come down.

And what we are seeing in terms of OEM pricing for SSDs, for example, 128 gigabyte SSDs are now roughly around the $25 level and 256 gigabyte SSDs are already down to about $45.

These SSD prices are very competitive compared to the $40 OEM pricing for HEDs.

So in a nutshell, while we're seeing strong demand when higher-capacity SSD prices converge with HEDs.

First, the 128 gigabytes, now 256, next 512 and later 1 terabyte.

And as each one of these SSD capacity price points converge with HED prices, we see big step-ups in incremental demand.

No, that makes a lot of sense.

And Wallace, I was wondering if you could kind of elaborate on the drop in Q3 for Shannon data center?

You talked about large hyperscaler having issues or changing with -- you said, new procurement timing.

I don't know really what that means.

I was wondering if you could elaborate on that.

And Riyadh, you're reaffirming the annual guidance.

Obviously, that implies a significant growth in Q4 both on a year-over-year basis and a quarter-by-quarter basis.

So on a year-over-year basis, it implies about 12% year-over-year growth.

So that means that growth rate is accelerating.

And on a quarterly basis, it's 9% sequential growth.

So I just want to get more comfort level on that business returning in Q4.

Well, as you know, sales of our Shannon SSD can be lumpy as individual project are generally very large.

Any small shift in timing can cause some big shift.

So project delay is not just because of controller solution, also could be procurement of [given] NAND.

So this up to customer.

I think the -- we do not believe our delayed Shannon sales are relate to any industrial trend.

Our delay is just project specific.

Our next question comes from the line of Mike Crawford from B. Riley FBR.

With this mix shift of the large SSDs from Shannon being softer in Q3 with less NAND pass-through the top line, what else potentially would hold expected gross margins below your 50% long-term target?

Well, as you know, we still have a lot of SSD solution sales.

Just because our SSD solutions are coming down sharply in Q3 from a big pushout of orders, it does not mean that we are not selling SSD solutions.

We still have a lot of SSD solutions coming from our Shannon products as well as our FerriSSDs, and all these products are below corporate average in terms of gross margin sales so blending down our overall gross margins.

Okay.

And then, given on the client SSD front, how you're working more closely with Intel, Micron and Western [Dig].

Are you seeing that your controllers are being paired more and more across the board with flash from those suppliers or -- and maybe less so with Hynix and/or Samsung in the past, or is it just always influx?

If we're working for Intel, we would be -- our controllers would be paired with Intel flash.

If we're working for Micron, Micron flash, WD with WD flash.

And for our eMMC, our eMMC controller paired with Hynix flash, right?

But if we're selling our controllers to module makers, they have a lot of discretion in terms of their sourcing of flash to be paired with our controllers.

That would be at their discretion.

Right.

So you're not seeing any change in kind of module maker choice flash that best pairs with your controllers?

We believe because our controller, the total solution already earlier than other merchant company controller maker.

So we are in a favorable position.

When the 3D NAND mature ramping with high volume, we take advantage [by all design] pipeline from both NAND makers as well as module makers.

Your next question comes from the line of Gokul Hariharan from JP Morgan.

First of all, Wallace and Riyadh, could you talk a little bit about how the dynamic of quicker adoption of UFS solutions affect your overall eMMC or -- overall smartphone-related controller business?

I think you still believe that your largest eMMC customer is unlikely to use your UFS solution once it migrates to UFS.

And could you give us some color about as this becomes more mainstream, are you going to under (inaudible) the smartphone market?

Are you going to be pretty much in line with the smartphone volume (inaudible)...

We believe our UFS controller will continue growth through the market transition from eMMC to UFS.

Although today, majority UFS is still in the flagship model, we believe from 2019 both MediaTek and Qualcomm, they will support UMCP for their mainstream (inaudible) processor.

So UMCP will start to be visible from first half of 2019.

And will be -- that will be the timing we'll see UFS will have a much faster growth rate and to replace eMCP and eMMC solution.

We believe Hynix already started UFS development for many years, more than 5 years.

So I think as today, it's very unlikely for them to adopt our solution for -- to their solution, however, there is no -- if it is possible for them to adopt us for the future product UFS.

Because if we develop earlier than their internal program, and they don't have a solution, they might also adopt our solution.

Our relation with Hynix is still very strong, very close.

But just -- I mean, different NAND maker have a different preference for internal or external third-party solution.

But we believe overall, our mobile storage controller will maintain stable, our eMMC might decline as the market decline, but UFS will grow as market grow.

So we feel comfortable and [welcome] our mobile storage controller solution.

Okay.

That's great.

Again, a couple of things on the open-channel controller.

I think we've seen some public announcements from Alibaba in terms of their adoption of open-channel, seem to have yet medium-term and long-term plan and seems like a pretty aggressive adoption curve.

I think -- last time, I think you guys have talked that we don't -- you don't have any revenue guidance at this point in time.

Can you talk about how the visibility is looking like, even that you started to have some revenues potentially at the end of this year itself?

And secondly, for some of these hyperscale partners in China, are you [this] sole partner, or are you one among a few controller solution partners?

Our relation with Alibaba is very, very close.

We have been codevelop the open-channel solution since a year ago.

So through the -- both the R&D team codevelopers specific, their software driver layer and our firmware layer and also even the hardware-specific serration for open-channel to meet Alibaba's different workload model.

And I believe this close relationship will generate significant revenue from 2019, expand to '20 and beyond.

And this will become one of China's standard for data center, not just Alibaba or BAT player.

We believe majority of e-commerce (inaudible) be private carmaker, they will also try to adopt open-channel because that can be customized to make specific workload meet their application need.

And Microsoft have their version of open-channel, and Microsoft also speaks strongly to make it happen in U.S. market.

And we believe similar to this model, maybe not just open-channel, maybe open-channel like a solution that were feeding other major player in U.S. and other countries.

So that make us SMI.

We have a -- very well positioned to use open-channel as a vehicle to penetrate a bigger hyperscaler and data center makers.

Okay.

So it sounds like you are the exclusive partner for Alibaba on open-channel.

We cannot comment on that.

Is it similar for your other hyperscaler as well?

We cannot comment on that.

We think we are the most close partner for Alibaba, and we also work with other hyperscaler together.

As you know, there is no single maker want to be exclusive only one supplier because that will be potential risk.

But we would like to [play to see] to have a major portion of the provider to Alibaba open-channel solution.

Our next question comes from the line of Mehdi Hosseini from SIG.

I apologize in advance if I'm asking a question that's already been asked.

I had difficulty dial in.

I want to go back to the push-out in hyperscaler.

I'm a little bit confused.

If this was a 1-quarter push-out, how come your December quarter is not (inaudible) strong?

And when I hear a push-out, I immediately think of the revenue that is pushed out from Q3 to Q4, but your overall year-end is still impacted by this push-out, which suggests we really don't know when that business is coming back.

So just trying to get more clarity on this.

And I have a follow-up.

In -- Q4 is typically seasonally weak for our Shannon products, and we were originally expecting weak seasonally for -- seasonality in Q4.

So the Q4 -- the Q3 sales has just been -- orders have been moved to Q4.

But at the same time, we are making up for that -- the lag because of a very strong SSD controller sales that we are now beginning to see.

Sure.

Let me rephrase the question.

Do you think this is a change in how that particular hyperscaler is planning or procuring for incremental capacity?

Or [has] there been some adjustment in how they procure from Silicon Motion?

Is it just an isolated case, or is that -- did it have a broad implication?

This is an isolated case.

We're expecting the sales come in Q4, and we'll -- we're expecting our projects -- we have multiple projects with our customer, and these will scale meaningfully into next year.

it's a still bit early for us to talk about potential seasonal patterns for next year, whether it will be like this year or like past years.

But off of the back of the programs that we have going in the second half of the year and where we'll ramp in Q4, we're expecting fairly strong growth from our Shannon products next year.

Got it.

Very clear.

And just a quick follow-up.

Going back to open-channel, should we still assume that material contribution is more of a second half of '19?

And would that also help with some incremental leverage in the P&L?

We're expecting our open-channel SSD products to go into production in the second half of the year once our last mile testing have been completed.

But when these products go into production, you should not expect large sales revenue in the second half of this year.

This is -- these products will go in -- when it goes production, the dollar revenue will be fairly small.

But based on the testing -- further testing, we're expecting pretty strong meaningful revenue from these products for next year.

But is that still like a -- second half of '19?

We believe, with our first half of 2019, the sale revenue will be very meaningful.

Okay.

And then any incremental leverage as these R&D -- as the past R&D investments turn into revenue?

We always have a lot of products.

This is one of our many products -- controller products, which we have devoted our growing R&D resources.

So this product, like all products, we're expecting strong returns from this investment.

(Operator Instructions) Our next question comes from the line of Charlie Chan from Morgan Stanley.

Wallace, Riyadh, so congratulations for great quarter and also longer (inaudible) SSD product cycle.

But for -- my first question is really again on enterprise SSD solutions delay.

I know it is what it is, right?

But what is the exact reason, right?

Because for Alibaba, first quarter is very busy, e-commerce season.

So did they give you any logical reason, meaning -- maybe they have sufficient capacity?

Or they thought NAND will be very available, so they are okay to delay.

Can you give us some colors on that?

[Chan], I think the -- to simplify the answer, it's really in the product transition.

As you know, we have [MPGA] product in the past from Shannon, although transitioning to the ASIC with both dual-mode open-channel and NVMe standard.

Through the product transition, we believe the original forecast is Q3 is pretty strong.

But through the product transition, we believe Alibaba have their preference and to see which product become more important for Ali, not just at 1 quarter, it more long term.

And everything for Ali is important for -- the real production test is for November 11.

So that is a milestone, they need to be 100% sure whether they want to focus on the legacy product or focus on the new coming transition for new product.

So this is their priority call.

And just for the product transition, and that's why see ramping for the new product delay for 1 quarter.

Okay.

Got it.

So it's nothing to do with (inaudible) CapEx (inaudible).

It's more like kind of the transition.

That is not.

So I think next week at Flash Memory Summit, you will know about how important open-channel will be.

Okay.

Yes.

And my second question is about your eMMC business.

I think it is sort of well discussed.

But if you may -- can you give us your assumption how much of eMMC (inaudible) today will return to internal next year for Hynix?

Charlie, the question is a little more complicated because we have the broader transition of eMMC to UFS.

EMMC technology for mobile devices is reaching -- has already reached the end of technology roadmap and is gradually beginning to transition to UFS.

For eMMC parts, as it relates to our business with Hynix, most of those -- especially the high-volume mainstream products will continue to be using our Silicon Motion controllers.

But gradually, these sales will also decline as broader technology shifts over to UFS, especially with what Wallace talked about earlier, the migration from the AP makers when they start supporting UMCP.

When that happens, we'll see more rapid migration to UFS.

And on the UFS side, we are making very strong headway into the marketplace with our partnership with Micron and have seen very strong growth in Q2.

And when these designs that our flash partner has secured with a lot of the flagship products.

When these rollout in high volumes in the second half, UFS sales will likely represent a meaningful part of our overall eMMC plus UFS sales.

So Charlie, let me just also add.

For eMMC transition for NAND maker, they will follow the market trend.

So even within Hynix, they will gradually transition when UFS portion was -- become bigger.

However, I think we also have the broader eMMC customer right now through the China local maker and China NAND maker, also including Unigroup.

And that will support certain -- from a broader playing, low-density and low-cost smartphone for India and China market.

So we feel very comfortable about our total diversified strategy.

And we believe our overall eMMC and plus UFS will become strong in balance, and then we'll very like to see the Micron is a strong growth for UFS.

And we believe SK Hynix portion eMMC will start to decline gradually, but we have other eMMC maker from YMTC and UNIC as our China and other maker continue to grow for low-cost eMMC.

All right.

So this year, you still hold your assumption that eMMC business can be flat.

And for next year, in the combination of all those smartphone storage, no matter eMMC or UFS, uMCP, do you think next year that part of business can keep flat?

Yes.

We are planning and managing our overall eMMC plus UFS sales as an ex growth flat revenue stream from here on.

There are no further questions at this time.

I would like to hand the conference back to today's presenters.

Please continue.

I would like to thank all of you for joining us today and your continuing interest in Silicon Motion.

We will be attending several conferences in Asia, Europe and the U.S. in the third quarter.

Details of this event will be available on our website.

Thank you, and goodbye for now.

Ladies and gentleman, that does conclude our conference for today.

Thank you for participating.

You may all disconnect.