慧榮科技 (SIMO) 2017 Q1 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the First Quarter Silicon Motion Technology Corporation's Q1 2017 Earnings Conference Call.

My name is Desmond and I'll be your conference moderator for today.

(Operator Instructions)

Before we begin today's conference, I've been asked to read the forward-looking statements.

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 limitations, statements regarding trends in the semiconductor industry and our future results of operations, financial conditions and business prospects.

Although such statements are based on our own informations and informations 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 and any change in our relations 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 obligations to update any forward-looking statements, which apply only as of the date of this conference call.

I would now like to hand the presentations over to our host, Mr. Jason Tsai, Senior Director of Investor Relations and Strategy.

Please proceed.

Thank you, and good morning, everyone.

Welcome to Silicon Motion's First Quarter 2017 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 some of our recent business developments, Riyadh will then discuss our first 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 the 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 the 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 on our earnings call.

I'm pleased to speaking with you again this quarter.

Our first quarter results was the high end of our expectation.

Our net revenue of $127 million and $0.70 in earnings per ADS represent a strong execution by our team, in spite of very tight NAND supply environment.

Riyadh will discuss our financial performance in greater detail later on this call.

As we had previously communicated, NAND supply has become very tight and this has been affecting our sales growth.

In the second half last year, our module maker customers were affected by NAND supply availability issues.

More recently, our NAND flash partners have also being affected by lack of sufficient incremental flash capacity from their own fabs.

These flash makers do not have sufficient NAND capacity to meet the need of their own products and need for their OEM customers.

Until large amount of new NAND capacity come online, NAND flash makers are all actively rationing their limited supply to partially fulfill greater demand.

As we had previously discussed, continued NAND flash tightness will trim the growth rate of our rapidly growing client SSD controller sales this year.

And separately, our sale of SSD solutions will decline this year.

We do not expect more significant amount for new NAND supply to come online until the second half of this year.

In our first quarter, sales declined due to seasonal factors and NAND flash tightness.

Both our SSD controllers and SSD solution sales declined, while our eMMC sales were unchanged quarter-over-quarter.

We believe that beginning in the second quarter, our sales will start picking up as our March pipeline of design wins will benefit from gradually increasing NAND supply and stronger seasonal demand.

So NAND flash industry has been adding a large amount of new 3D capacity, since the second half of last year.

The CapEx spend for 3D NAND is one of the large CapEx spends that we have seen in many years.

Other than Hynix, all of the flash maker have already begun production of 64-layer 3D NAND, and they are all currently aggressively tuning production yield.

We are confident that we should start seeing meaningful 64 layers 3D supply in the second half of this year, as yield are now approaching level that are suitable for mass production.

And in the second half of this year, we expect Hynix to start to gradually ramp their production of 72-layer 3D flash.

During next year, you should expect to start seeing initial production of 96 layers as well as 4 b/cell QLC 3D NAND.

One of the biggest build out of NAND supply is beginning to come online, and we should start benefiting in the second half of this year.

This current and upcoming generation of the NAND flash technology will continue to drive down NAND dollar per gigabyte cost, which should lead to continue SSD growth, proliferation of the NAND-based solid state storage devices and the demand for our controllers.

Let me now move down and talk about our key products.

We continue to see strong demand for client SSD from PC OEM and consumers, especially for commercial notebook PC, despite current NAND supply tightness and resulting high NAND and high SSD prices.

Many new notebook PC are designed with SSD for reasons relating performance and some factors.

Using HDD as an alternative is no longer possible.

Leading third-party research firms like Gartner are now recommending to their corporate customers that all PC procurements should be using SSD, therefore, both notebooks and desktops, because of better performance, reliability, power consumption and total cost of ownerships.

We are confident that, over the next few years, as the cost of the NAND continue to decline (inaudible), the majority of 500 million a year storage devices market will be served by SSD, as HDD are displaced.

Our industry-leading highly customized -- customizable turnkey hardware plus software client decontroller sales decreased in the first quarter, as expected due to seasonality.

Despite these challenges, our clients' decontroller sales grew nearly 40% in the first quarter as compared to a year ago.

And we remain on track to grow our clients' decontroller sales 20% to 25% for the full year.

We remain the leading merchant controller supplier for client SSD, with a market share of approximately 30%.

Through our flash maker partners and module maker customers, we continue to maintain a strong presence in both the PC OEM and channel markets.

We're starting year with more clients decontroller design wins than at the same time last year and with a much broader portfolio of controller solutions from SATA 3 to PCIe NVMe, that range from high performance application to low cost DRAM-less and PCIe PGA devices.

We believe sales on our pipeline of clients' decontroller design wins will scale when NAND tight supply improve in the second half of this year.

Almost all our new design wins that will enter production this year are for PCIe NVMe SSD and will use either 48- or 64-layer 3D NAND.

Let me update you on our the enterprise grade decontroller development.

As we discussed last quarter, we are currently developing turnkey enterprise-grade controller solution.

More specifically, we are currently engaged and in active product development for an OEM on several enterprise-grade turnkey SSD controller projects that involve both engineering of both highly customized enterprise-grade controller, hardware and software.

These SSD controller projects are all for hyperscale data center, and I expect it to go into production next year.

These SSD with our controller are performance optimized and designed to deliver performance superior to standard off-the-rack high-end enterprise-grade SSD.

Overall, we are pleased with our delivery of our SSD controller targets and growth objectives, our product road maps expanding portfolio for design wins, coupled with factory material improvement in supply availability beginning in the second half this year, point to continuing strong SSD controller sales growth next year.

Let me now turn to our eMMC and UFS controller.

Demand of our eMMC controller remained steady in the first quarter, in line with both our expectation and the past seasonal patterns.

Demand for our SK Hynix remain stable and demand for eMMC and eMCP embedding memory for smartphone OEMs in China and elsewhere, also remain stable.

We are also seeing broadening adoption of eMMC beyond smartphones and tablets, smart TV, streaming set-top box, Nintendo Switch console, (inaudible), Amazon Echo speakers, automotives and other AI and IoT devices.

Our UFS projects are progressing as expected.

Our NAND flash partners using our UFS 2.1 controller has already secured its first discrete (inaudible) embedded memory designing with the global top 5 Android smartphone OEM and is on track to begin volume production in the second half of this year.

Separately, we expect our UFS controller for SK Hynix, UMCP to launch later this year.

But it is increasing likely that timing of UMCP support by MediaTek and other leading AP vendors will be delayed until the first half of 2018.

We are very well-positioned for this eventual high volume rollout of UFS.

But until that happens, revenue contribution from UFS sales will remain small.

We remain comfortable that our eMMC and UFS controller sales this year are on track to growing above 5%.

Finally, I will discuss our SD solutions.

Demand of our Shannon enterprise SSD continue to be very strong and new projects with our primary Chinese hyperscale customer has already starting shipping in the second quarter.

The pace of the infrastructure build out is accelerating, driving strong incremental demand for enterprise SSD.

If not for the tightness in the NAND market we are all facing today and the last price supply commitment that our customer has being able to secure, we believe our Shannon enterprise SSD will have grown this year.

Sale of our Ferri industry SSD, unlike our Shannon enterprise SSD, were stable in our first quarter and will also grow in the second quarter.

To summarize, despite our growing pipeline of business with customers and expanding portfolio of controller and SSD solutions, our business momentum this year, especially, in the first half this year has dampened by NAND industry supply tightness.

This headwind to our business should lessen in the second half of this year, when new NAND supply come online, and we should be very well-positioned for 2018 growth.

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

Thank you, Wallace and hello, everyone.

I will summarize our financial results and outlook.

Before I begin, I'd like to reiterate that our comments today will focus primarily on non-GAAP results, unless otherwise specifically noted.

A reconciliation of our GAAP to non-GAAP data is included with earnings release issued today.

Our first quarter revenue decreased 12% sequentially, but increased 13% as compared to the first quarter of last year.

For the second quarter, we are expecting revenue to increase 5% to 10% sequentially, driven by growth from our embedded storage products, which are our client SSD and eMMC controllers, plus our industrial and hyperscale SSD solutions.

Our embedded storage product sales continue to be roughly 80% of total revenue in the first quarter and will remain roughly 80% of total revenue in the second quarter.

Let me provide additional color on our key products.

Our client SSD controller sales declined by approximately 20% sequentially, but increased about 40% year-over-year.

Sales continue to track towards 20% to 25% full year growth for 2017 with stronger sales expected in the second half of the year.

You should expect our client SSD controller sales to become materially bigger than our eMMC for full year 2017.

eMMC controller sales growth were approximately flat sequentially, but had increased about 15% year-over-year.

Sales continue to track towards approximately 5% full year growth for 2017, with second half sales comparable to the first half.

SSD solutions declined approximately 10% sequentially, but decreased about 20% year-over-year.

Sales of our SSD solutions are still tracking towards a full year decline of 15% to 20% in 2017.

Our Ferri industrial SSD should grow this year while our Shannon hyperscale SSDs will decline.

Our first quarter gross margin increased to 51% from 50.2% in the previous quarter due to temporary lower cost of sales relating to our SSD solutions.

Our first quarter SSD solution sales were utilizing lower cost NAND inventory procured a few months ago.

This temporary benefit will reverse in the second quarter as more recent procurements of NAND have been at a higher cost.

This is why we are expecting our second quarter of gross margin to revert to the 48.5% to 50.5% range.

We value our inventory based on average cost method.

For the full year, we continue to expect gross margin in the 49% to 51% range.

Our first quarter operating margin decreased to 26.4% from 30.4% in the previous quarter due to higher operating expenses and lower revenue, offsetting the benefit of higher gross margins.

Our first quarter operating expense of $31 million is the same amount as the operating expense implied by the midpoint of our previous revenue, gross margin and operating margin guidance.

Our first quarter operating expense is 10% higher than the previous quarter, primarily because of the normalization of reoccurring accruals and foreign exchange changes.

For most of last year, we over accrued a certain compensation expense and reversed this accrual in the fourth quarter, and this resulted in the first quarter's accrual being compared against the reversal.

Additionally, roughly 2/3 of our total employees are based in Taiwan and paid in NT dollar terms -- in NT dollars.

And in the first quarter, the NT dollar strengthened sharply against of the U.S. dollar resulting in much higher compensation expenses in U.S. dollar terms.

Our total headcount this quarter increased to 1,149, which was 27 more than the fourth quarter.

Tape-out expenses were similar to the previous quarter.

Looking into the second quarter.

We are expecting operating margin in the 25.5% to 27.5% range, with midpoint of this range similar to our first quarter operating margin.

We are expecting operating expense next quarter to be similar to our first quarter.

Our effective tax rate in the first quarter was 27%.

This is higher than the 20% model tax rate.

We anticipate our effective tax rate to remain elevated for 1 or 2 more quarters due to the mix of operating entities that are profitable and incurring tax expenses with entities generating pretax losses.

Our first quarter EPS was $0.70, lower than last quarter's $0.95.

Total stock-based compensation in this quarter was $3 million, lower than the $8 million in the previous quarter due to the seasonal timing of RSU awards and consistent with the timing of past years.

For the full year, we expect stock-based compensation in the range of $15.5 to $16.5 million.

Let me now talk about a few key items from our balance sheet.

We had $304 million of cash, cash equivalent and short-term investments, $27 million more than in the previous quarter and $113 million more than a year ago.

In February 2016, we paid $7 million of dividends to our shareholders, second $0.20 per ADS quarterly installment of our annual $0.80 per ADS dividend.

This concludes our prepared remarks.

We will now open the call to your questions?

(Operator Instructions) Our first questions comes from the line of Mike Burton from Longbow Research.

Just I wanted to ask, first, on the eMMC, the mobile business.

Curious what you are seeing there?

I realize, you ship in a little bit earlier into the flash OEMs, but clearly many suppliers have been affected by an inventory correction in Chinese handsets.

Given your orders, what is -- what's the outlook there versus normal seasonal patterns for the second half?

We believe we see the second quarters will be in line with the first quarter.

Second half, it depend on the inventory adjustment.

Current forecast, we believe, is similar to the first half.

Okay.

And then also just following up there too on the UFS side.

It sounds like MediaTek support certainly has pushed out adoption.

What's your expectation for UFS as -- share versus eMMC for the market as a whole this year and next?

And can you help us understand how you plan to win versus some of the internally supply controllers?

The currently major UFS adopter is Samsung Mobile, as all the Galaxy series use adopted UFS solution, and we do see this year, there would be more say China smartphone maker, the flash model adopt UFS.

So we believe this year, the UFS market share will be around about 15%, 20% among the Android smartphone.

But however, resulting UMCP will be very difficult to UFS to become the mainstream, because majority smartphone maker will like to combine NAND and DRAM together into one PGA package.

We believe this renewed effort from MediaTek to become main player for -- in mainstream smartphone chipset to adopt UMCP maybe in second half 2018 or 2019.

Okay.

And then, I apologize if I missed it.

But any update on the 4th flash OEM for client SSD?

I realize, it's not in the guidance, but does it look closer to ramping this year versus next?

And any color you guys can help with on what are some of the factors holding up on that ramp?

I think the timing of this new project is dependent on the improvement -- improving availability of the NAND.

With the NAND condition remaining tight, our 4th flash partner has postponed the rollout of this new product until availability improve, but that's the current situation.

Next question comes from the line of Suji Desilva from Roth capital.

So you've reiterated the full year revenue guidance, but the second quarter was a little bit below where I was, so it makes it a more back-end loaded year which is consistent with your commentary.

Would -- in this scenario, would the fourth quarter be above seasonal and actually show an improvement from what you expect?

Or will be very third quarter loaded?

So from our perspective, there is not much change in terms of our full year outlook and also the sequential quarterly pieces that adds up to their full year numbers.

Our first quarter numbers are little higher than or in the high end of our original guidance.

Our second quarter numbers are pretty much consistent with what we were originally expecting, and so second half of the year is going to be second half loaded for a full year.

So we're expecting our business to grow each quarter through 2017, as NAND availability gradually improves this year with stronger growth in the second half.

If NAND supply does increase faster than what we're currently expecting, we can certainly see some upside to our current expectations.

But conversely, if the challenge is in improving manufacturing yields of 3D NAND are not resolved as expected, we could see potentially tighter NAND flash availability.

But on balance, this scenario doesn't seem -- seems less likely.

So we're continuing to expect that the tightness in the market will be resolved later this year, and we expect our 2018 growth to strengthen materially from this year as all the NAND flash makers ramp their 64- and 72-layer 3D into full production.

Let me add this point.

The reason we have confidence for Q4 which should overcome the seasonality, because through the current design pipeline, as the NAND supply is sufficient, we're confident that Q4 and second half will have much stronger sales growth.

Okay, great.

That's very helpful color.

And then also on the SSD with the NAND tightness here, can you tell us where the PC attach rates are in your perspective?

And where you think they'll be exiting '17 and perhaps in '18 and whether that's kind of below your prior expectation, given the tightness or tracking to it?

So Suji, PC pass rates should be at a higher level, but as you know, NAND flash availability is affecting the amount of NAND that can be used to build client SSDs and also the current higher price of NAND and higher price of SSDs is also impacting the transition of ACDs to SSDs.

So those are factors that are affecting where we are now.

But we should expect to grow around 20% with client SSD average.

Next question comes from the line of Rajvindra Gill of Needham & Co.

So Wallace, a question on the enterprise-grade SSD turnkey solution targeting hyperscale customers, you're saying next year production.

Can you maybe elaborate in terms of what the design process has been like?

What the transition has been from a technological perspective from moving from client SSDs to more enterprise-grade SSDs?

And how your sales force has adapted and your engineers have adapted to that transition?

And we see from late last year this year, the strong demand for data center, and I think that's one of the main reason causing the NAND tight supply in 2017, and we see the strong demand not just North America, China, it's through all the region worldwide.

And we see there is a tremendous demand regarding enterprise SSD solution as well as controller, because all the customers looking for differentiable solution.

Clearly, I think from last year Q4, we made a determination to develop enterprise-grade SSD controllers.

Currently, we're working closely with 1 leading and major worldwide enterprise customers, and we do work closely.

We have a face-to-face meeting every month, weekly conference every week to review the schedule and all the key feature and focus on the production in early 2018.

And enterprise customers, they do need standardized design NVMe and SATA solution.

They also need a very high performance and suitable for this (inaudible) data center.

So that is very important for them to be more flexible to use a solution can serve these end purposes.

So that's why we tailor our controller, both hardware and software, to meet enterprise customer need.

We brings the solution to bring our SSD portfolio to the next level and can quickly help us to grow both controller business and Shannon SSD business.

Very good.

And on the client SSD side, kind of maintaining the 20% growth year-over-year on average, can you perhaps talk about the competitive landscape?

I know one of the other competitors in the space have ceded a lot of share to you, but they're now saying, they have the turnkey solution with firmware and software -- firmware and hardware.

Wondering if you're seeing any competitive pressures on the client SSD side?

We cannot comment on our competitor.

But we can say, we are still the leading merchant supplier of turnkey controller solutions for the client SSD, for mainstream, high volume SSD.

Our flash partner and module maker customer prefer turnkey solution now being able then to gain market faster and more cost efficiency.

I think, we offer from SATA 3 to now PCIe NVMe solution.

Our second-generation PCIe NVMe just coming out.

We offer Gen3 x4 A channel, Gen3 x4 4 channel, PGA, PCIe, PGA solution as well as DRAM-less solution.

So we have a stronger, broader portfolio with our turnkey solutions.

We do not see any meaningful attraction from any other emerging controller supplier.

We act as an extension of our flash partners OR&D and enable them to go after broader range of the end market quickly and cost efficiently.

Currently because of tight NAND supply, every NAND maker have certain preferred allocation for NAND.

But we believe, our turnkey solution will continue to be preferred, and we believe we will continue to grow with our clients SSD.

And currently in Intel, Micron, WD are using our turnkey controller solution.

And we mentioned, the 4th NAND maker when NAND flash supply is sufficient, and we'll move net production.

And last question from me.

You'd mentioned that, overall, the production yields on 3D NAND for the rest of the other flash makers are going up and they are aggressively tuning the yields.

Can you maybe provide some color in terms of kind of what percentage of overall NAND supply you think is 3D NAND, currently?

And then as you mentioned SK Hynix is gradually ramping the 72-layer and that just start to positively affect your business.

Can you talk a little bit about their process?

And how the transition to 96-layer next year is going to affect your business positively?

I think, we cannot comment regarding the NAND makers, their own strategy and mix.

As you know, the transition from 2D to 3D really take time.

All flash maker, they are more cautious to invest the capital equipment before they see visible and meaningful yields for transition.

So the transition is very dynamic.

And but the good news is, we see all the NAND maker ready to ramp the 64-stacked 3D NAND.

And Hynix also will be ready to ramp their 72-stacked 3D NAND following Q3 this year.

As to how soon, how fast they can ramp into the scale and -- we cannot comment.

But we believe, NAND supply will be improving in the second half, and we believe the NAND supply will be more sufficient in 2018.

That will be our benefit due to SMI.

Regarding 96-stacked 3D NAND and the quad b/cell QLC 3D NAND, and we already discussed in case with the other NAND maker, we'll provide a solution.

Our controller firmware will be ready to support the QLC 3D NAND and when they move to mass production.

Next questions comes from the line of Anthony Stoss from Craig-Hallum.

Wallace, maybe if you won't mind commenting, you mentioned that the bulk of your SSD wins for the second half year are all PCIe.

Refresh our memory, the bump you see in ASP and if that's changed from what you expect in the past.

And similar question on the enterprise server market, as you launch in early '18, maybe a range what you would expect on enterprise server controller would be helpful?

Tony, I'll take this call -- question.

Our ASPs as you know, for all of our new products, they always come in slightly higher than our corporate average.

But as these products become mainstream, then our ASPs converge towards our average ASP for the specific class of products.

So this is going to happen again with -- and is happening now with our new PCIe products.

New products coming in higher ASPs and as they ramp and they become mainstreams, our ASPs will converge.

The question about enterprise-grade control ASPs, this is proprietary and it's not an arrangement we're going to be talking about on calls.

Okay.

Follow-up question then, overall in gross margin, given your steadily increasing SSD revenue and the layering on of enterprise, would you expect your kind of the next move, your 49% to 51% gross margin that will go as higher some time in '18?

Or how should we think about gross margins longer-term?

We have a business model, as you know, that has a wide range of products.

Certain products, the gross margins are higher than corporate average, in other products, the gross margin are lower.

And we actively manage our -- the components of our products with higher and lower gross margin, and we've been able to deliver our gross profitability at roughly 50%.

And we'll continue to be able to deliver gross profitability towards the 50% going forwards.

And you should expect us to manage our business with this objective in mind.

Next question comes from the line of Mehdi Hosseini from SIG.

I want to go back to the MediaTek commentary and better understand your overall market share, especially, the Android.

It's very obvious that MediaTek may have lost market share.

And I'm just wondering if you have as much exposure to the other AP providers as you do to the MediaTek architecture?

Or any kind of commentary that you can provide would be helpful as how the market share dynamic or competitive landscape in the AP market is impacting you?

And I have a couple of follow-ups.

I apologize.

Maybe I did not state regarding MediaTek AP clearly.

Actually our eMMC, eMCP can work for every chipset maker, including Qualcomm, MediaTek, Samsung, Intel or Spectrum.

And it has nothing to do just with MediaTek only.

What we mentioned is, our UFS, UMCP product, which is -- it will allow MediaTek to promote because their chip size is more suitable for UMCP for mainstream market and it looks like this year, they're going to delay the support for UMCP.

That's why we believe that production for our UMCP maybe would delay from late this year to first half of next year.

Doesn't that imply that your UFS share maybe less than prior generation primarily due to the fact that MediaTek is late?

No.

It's -- UFS today is still very small between 15% to 20% for overall market, and Samsung Galaxy is the primary adopter for UFS.

But still remains very small portion of our total Android smartphone.

We maintain about 30% to 35% market share.

We saw eMMC and eMCP, and SK Hynix is our major customer for eMMC controller.

Okay.

And then on your exposure to the Chinese IDC, Internet Data Center, service providers, it seems to me that something has changed, because in the past you were of the opinion that those kind of customers would have the scale to be able to procure NAND well in advance.

And now it seems like the NAND shortage has adversely impacted repeat business.

I'm just trying to understand to what extent it is really -- is it in fact NAND shortage?

Or whether it has more to do with design cycle?

And what those customers are doing with their products?

It seems to me that there are 2 separate issues going on.

As we mentioned, this year, specially the NAND shortage start from, see, June last year.

But last year the shortage was due to the smartphone doubled the density of the storage, not just iPhone, all the Android phone doubled density.

That momentum caused the NAND shortage.

This year, it's quite different.

The main reason causing NAND shortage is infrastructure change, strong demand from data center.

So it's not just a storage point, it's indeed the tape drive, the HDD, all increased the density, due to the strong demands on data center.

Regarding the -- our specific customers, they want more SSD solution to expand their infrastructure, because the mobile user per terabyte has increased very, very quickly.

Every Internet e-commerce company want to solve data traffic problem.

They want the faster, load latency.

I think it's not China maker, all the leading U.S. company, they also suffered NAND shortage for this year.

I think, there is a global issue, is very common issue, everyone known, it's not particular for one customer and this is the whole NAND industry the NAND is seeing severe shortage and tight supply.

Sure.

And then one quick follow-up for Riyadh.

Your year-end revenue guide of flat to up 10%, does the up 10% imply that the 3D NAND yields have improved and there is no shortage?

Or does that mean that there is just some marginal improvement, so that if the yields were to improve at the faster rate, there would actually be upside to the high-end of your revenue guide?

Our guidance is more the latter of what you stated.

Since our customers are still working through their -- our business partners, our flash partners are still working through their production yields, they really are not in a position today to be willing to commit to volume that they still have not been able to deliver, because they are still improving the yields.

While they have planned yield improvement plans, they still need to achieve those numbers.

And until the achieve those numbers, those numbers are still a planning number.

And so going back to our numbers, we're working more on the conservative set of numbers without the high yields being rolled into expectations.

So if the flash makers are able to achieve the higher yields, there potentially could be upside to our numbers, but we currently are not able to commitment -- to commit to that, because our business partners haven't achieved them yet, and we're dependent on them to achieve before those numbers roll into our numbers.

Next question comes from the line of Charlie Chan of Morgan Stanley.

So I wanted to follow up the question regarding storage density, because Wallace just mentioned that last year, China smartphone doubled their density.

This year, we see that NAND cost is doubling.

So do you see that both for eMMC and SSD, end customers are downgrading the NAND density per dot to fulfill more devices with those SSD feature?

We believe the smartphone maker, they also get a vibe from NAND supplier, because they do see the shortage.

So we see the densities will remain the same as the second half of last year.

This year major is due to the strong demand from data center and much stronger than everybody planned early this year.

And even our major customer from data center, or Internet e-commerce company, they also supply regarding the stronger demand globally.

And we do -- we cannot comment regarding iPhone, other's plan.

But for general speaking, for all the Android Smartphone density maintain the same as the second half last year.

Charlie, let me also add, there is a bifurcation on the client SSD side.

For the client SSDs going to for PC OEMs, it seems that the densities have been inching up.

But on the other hand, for client SSDs going to channel market as flash availability is highly rationed, a lot of the channel players have been building at lower density.

So there seems to be that bifurcation because of higher prices.

And similarly, if for the channel side of market because of the high prices, it looks like they may be able to be -- able to do business more profitably by going after the lower density -- densities.

Okay.

Got it.

And also on the 3D NAND, so we all know that your partners are building up production et cetera.

But does that purely absolute efficient efforts (inaudible) overall NAND flash in third quarter efficiently and grow over second quarter?

They're absolute efficient numbers.

Yes.

It looks like that way.

We're expecting to see bit shipment growth to accelerate in the second half of the year, as a lot of the new 64-layer capacity reach go into full production.

Going to full production still means that there's still lot of yield improvement upsides to be delivered.

But the full production should be happening more in third quarter second half time frame and when that happens, we should start seeing more significant quarter-over-quarter bit growth compared to what we're seeing in the first of the year.

Got it.

This is what I wanted to know.

And lastly, on the PCIe -- sorry, the enterprise SSD control IC business, you want to run it early next year, right?

So in terms of (inaudible), do you think majority of that will be PCIe or maybe you have PCIe Intel's SATA?

I think the enterprise server or data center all have different demand.

Today SATA still maintain the large percentage also SaaS, that's still growing to replace SaaS HDD.

But the PCIe definitely has the stronger momentum.

And to our product, we're focused on PCIe for enterprise, and we will offer both the rack I/O, hybrid I/O.

So we have different software matching to meet the customer need, tailored for the customers' requirement.

Next questions comes from the line of Tom Sepenzis from Northland.

I'm just wondering given the expected growth in NAND, when should we expect that pricing will come down enough for you to start seeing some movement on the Shannon business?

It's a very -- actually difficult question.

When NAND supply becomes improved, we are supposed to see the price come down.

So I believe NAND maker may have different pricing strategy and supply strategy.

So we believe the NAND price will maintain stable or little higher in Q2 and moving into Q3.

But from late Q3, maybe the price will become soft when there is more supply coming in the market.

So should we expect that business to pick back up in 2018?

Definitely.

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

In the past month, we've seen SSDs come out with more favorable reviews, the SSDs including your SM2260 controller.

So I'm wondering if you think that is true versus -- those reviews are accurate versus more earlier reviews and how often you update firmware for that controller?

Or how much of that might be due to more stable NAND from your OEM partners?

I think some reviews where -- that are true, because those are very early stage and fresh product.

We continue to improve our firmware and upgrade to our customers.

And because today due to the tight NAND supply, so if the customer have lack momentum to launching a new product.

But our second-generation PCIe are coming very strong.

We believe and many, many customers can see our product coming around late Q3 into the market.

And we offer, as I said, offer PCIe Gen 3 x4 A channel and 4 channel for mainstream.

They are all designed by 28 nanometer process and with the very high performance, low power and can -- very competitive and the leading player today.

So you would attribute that better performance to more to your second-generation controller than to more stable NAND flash?

As you know, the SSD performance it depend on controller hardware feature and the firmware algorithm as well the NAND itself, and so they all combine together and also depend on density, right?

So we cannot comment who plays the most important role.

But we believe, all the issue will be addressed by our second- generation PCIe product line, which have much stronger positioning because they can interface with all type of different 3D NAND, doesn't matter TLC or QLC.

Okay.

And then last question is, on these enterprise-grade controller projects you're working on, when you started generating sales in these controllers?

What type of ASP are we looking at?

And will this -- will these be grouped in with Shannon sales or else-- somewhere else?

Mike, as I previously talked about, we're not going to be commenting specifically about the pricing -- the prices of our enterprise-grade SSD controllers, given the proprietary nature of our business arrangement with our customers.

But as we talked about early in the call, we got several of these hardware plus firmware SSD control projects with this OEM for hyperscale data centers and these projects are going to enter into production in 2018 now.

Wallace had also talked more about this.

But let me also add, even before this project, we were already part of enterprise SSD market.

We already -- we've been supplying our highly customized Shannon hyperscale SSD solutions and our -- and then we've been also supplying our controllers client-grade SSD hardware with especially developed enterprise-grade firmware.

And so this is the continuation of what we're doing.

Let me add a comment.

Regarding the business model, with enterprise solution under Shannon operation, we focus on one particular OEM and after went to mass production, we have 2 other major OEMs we engage.

So regarding enterprise controller, they'll be under operate by SMI.

We have a follow-up questions from Rajvindra Gill from Needham & Co.

Just to read out a follow-up on the tax rate.

So should we be modeling kind of 27% for second quarter and third quarter and then have that come down back to 20%?

Or how should we think about that?

That's correct.

We're -- our tax rates are little elevated for the first quarter and will remain elevated for 1 or 2 quarters more given that we have a mix of operating entities that have higher -- that are profitable and entities that are making pretax losses, which is dragging down our overall pretax income.

And so when you sort of look at our effective tax rate calculated on a consolidated basis, that's going to pull up our effective tax rate.

But longer-term, you should expect our tax rates to normalize and we should be back down to 18% next year.

(Operator Instructions) There are no more questions from the line.

I will like to hand the call back to Mr. Wallace Kou, CEO, for closing remarks.

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

We will be attending over a dozen conferences in Asia, U.S. and London this quarter.

Details of this event will be available on our website.

Thank you, and goodbye for now.

Ladies and gentlemen, that does conclude the conference for today.

You may now disconnect the line.