慧榮科技 (SIMO) 2019 Q2 法說會逐字稿

Ladies and gentlemen, thank you for standing by, and welcome to the Silicon Motion Technology Corporation Second Quarter 2019 Earnings Conference Call.

(Operator Instructions) I must advise you that this conference is being recorded today, Wednesday, 31st of July, 2019.

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 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 and 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.

I would now like to hand the conference over to your first speaker today, Chris Chaney, Director of Investor Relations and Strategy.

Thank you.

Please go ahead.

Thank you, Alyssa.

Good morning, everyone, and welcome to Silicon Motion's Second Quarter 2019 Financial Results Conference Call and Webcast.

My name is Chris Chaney, Director of Investor Relations.

With me today is 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 second quarter results and our outlook.

We'll 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 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 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.

And with that, I would like to now turn the call over to Wallace.

Thank you, Chris.

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

First, some financial highlights before I talk about our business.

Our second quarter sales excluding FCI increased 6% sequentially to $94 million.

Earnings per ADS for the quarter were $0.52, up from $0.42 in the first quarter.

Let me start by updating everyone on our business starting with our SSD controller product line.

In the second quarter, SSD controller sales grew sequentially at a solid 15% growth rate.

Our SSD controller sales to NAND flash makers grew 30% sequentially.

Our SSD controller sales to module maker customers, on the other hand, were softer than anticipated as they continue to be risk adverse because of rapidly falling NAND prices.

With the NAND prices beginning to stabilize, we believe our activity with module makers will pick up.

For full year 2019, we believe our SSD controller unit shipments should grow at least as fast as the growth rate estimated by industry analysts.

Gartner for example is projecting at least 20% climb in SSD growth.

In terms of our customers and end markets for the full year, we expect stronger sales growth from both our NAND flash customers and module maker customers.

And we expect stronger sales growth to the OEM market relative to the channel market.

Our customers supplying to OEMs are benefitting as they are seeing very strong SSD design activity for PC, game consoles, and other client devices.

This year as OEMs focus on SSD and transitioning their products from HDD to SSD, PC shipping with SSD will grow rapidly, and conversely, PCs using HDD could fall by as much as 50%.

Our SSD controller growth this year is being driven entirely by PCIe, eMMC SSD as sales of SATA SSDs decline.

OEMs are almost singularly focused on eMMC SSD and all of our new controller design wins for OEMs are for NVMe SSD.

Channel markets are primarily SATA SSD, but are also gradually beginning to transition to NVMe.

We expect to exit the year with more NVMe controller sales than SATA sales.

Our SSD controller design activities continue to expand and our market share likely also increased in the second quarter.

We continue to out ship our primary US competitor in terms of NVMe controllers and have secured a significant majority of design wins awarded by NAND flash makers to merchant controller suppliers.

We are also seeing our smart Taiwanese competitors lose traction in SATAs with customers due to their more limited capabilities.

Secondly, we do not believe our NAND customers are losing market share to NAND vendors who are not our customers.

And more broadly, we are seeing increasing business opportunity because growing interest by NAND vendors to outsource SSD controllers, both their mainstream and entry level client SSD.

We believe we supply at least 3/4 of controllers procured by module makers in Taiwan and China.

And globally, we supply roughly 1/3 of the overall market.

Our client controller business however is not without challenges.

In the past, our planned SSD controller prices have been very stable.

In the near-term however, we expect blended prices to drift down because of 3 trends.

First, our customers are increasingly focused on cost versus performance in managing their SSD product portfolio because SSD performance is already much better than HDD.

Lower cost product improves mainstream and entry level SSD and these devices use lower cost controllers.

For example, if a customer migrates from SATA to PCIe NVMe, they also are shifting a bigger portion of their requirements for more expensive SSD that will require D-RAM to less expensive DRAM solutions.

A related technology trend is as capacity per NAND die increase faster than SSD capacity, SSD manufacturers are using 2 die per SSD and are also turning to lower cost, fewer channel controllers.

Such as 2-channel versus 4 channels or 4-channel versus 8 channels.

The other industry trend is rapidly falling NAND prices have been negatively affecting the profitability of both NAND flash makers and module makers.

It is in our interest for our customer to survive and for us to maintain our diversified customer base.

It is also in our NAND side partners' interest that a diversified module maker customer base is maintained so they benefit from a broader customer ecosystem.

To maintain this diversified partnership ecosystem and since we supply to almost all these module makers, we need to be flexible with our controller pricing during periods of challenging industry conditions.

Industry consolidation is not in our interest.

The other reason our blended ASPs are currently lower than anticipated is we want to price out our competitors seeking a small foothold in our marketplace.

We believe we have been successful with this strategy.

Let me turn to our enterprise SSD controllers.

These controllers are now shipping, first to customers in China and in the fourth quarter to U.S. customers.

Our enterprise SSD controllers are now using our open channel SSD that are now commercialized, commercially deployed in the data centers of Alibaba and other B-A-T customers.

In the fourth quarter we will begin initial sales of these enterprise controllers to a U.S. customer for their data center standard NVMe SSD.

Sales of our enterprise SSD controllers this year will be small, but we are encouraged by these initial milestones and expect more meaningful contribution next year.

I look forward to sharing more about our enterprise SSD controller business activity in upcoming quarters.

Turning to our eMMC+

UFS controllers, while we have not seen any meaningful pickup in Android smartphone build activity especially in China, our eMMC+

UFS sales this quarter were much better than expected, growing 20% sequentially.

This upside is a result of our Chinese module maker customers bringing their eMMC embedded memory solution to Chinese OEMs ahead of expectations.

As we had previously discussed, we have been actively supporting 6 to 7 module makers entry into eMMC market for indications ranging from smartphone to set-top box and speakers.

We expect our eMMC controller sales to these customers to scale rapidly in the next few quarters.

Our large NAND flash eMMC customers continue to diversify their overall NAND sales away from the mobile market.

Nevertheless, we continue to expect our sales of eMMC controllers to this customer to remain stable for the rest of the year.

For other NAND customers' UFS programs, we continue to make good progress in helping them bring their product to their OEM customers.

Recently we helped this partner launch a second-generation UFS solution with best in class endurance.

Now I will provide some detail on SSD solutions.

In the second quarter, SSD solution sales fell about 40% sequentially, with our Shannon product accounting for most of the decline.

Orders for our legacy Shannon product declined sharply and missed internal sales targets.

Our sales of [Victera] self-defined storage appliance and Ferri product also missed expectations.

More positively, our open-channel NVMe SSD are now commercially deployed in the data centers of Alibaba and in other B-A-T customers.

These are important milestones for us and also for the SSD industry as these open channel SSDs are the first commercial deployment of this technology in the world.

Applications operations from this datacenter are now using our Shannon open-channel SSD which enables SSD performance to be optimized for applications, a benefit not available with standard SSD.

While these open-channel SSD milestones are very important for us, both of our B-A-T customers have significantly reduced procurement plans of our product in the second half this year.

With these SSD continuing to perform satisfactorily, we are well positioned to bring our second generation open-channel SSD for qualification and testing at Alibaba and the other B-A-T for the next year projects.

We acquired Shannon on May 2015 to leverage our unique controller technology to create incremental value by supplying customized enterprise SSD.

Analogous to our customized (inaudible) SSD and also as a vehicle to accelerate our enterprise SSD controller development.

Also, we cannot just sell enterprise SSD controller to Chinese hyper-scalers.

We need to provide them with a complete SSD solution which embeds our controllers.

2019 is also a challenging year for our other 2 SSD solution products.

We customized [VJ] form factor Ferri (inaudible) and our new Victera software defined startup appliances which enable virtualization of enterprise storage infrastructure.

Sales of our Ferri SSD declined sequentially as NAND prices have fallen sharper than expectation.

We are excited about our long-term Ferri growth giving our growing and diversified portfolio of automotive infotainment, commercial printer systems and regulated gaming equipment design wins.

We believe the worst is now behind us and looking to next year we believe we are well positioned for renewed growth.

Our SSD controllers are expected to continue growing.

Our eMMC+

UFS controllers are expected to be flat to up next year.

2019 will be very challenging year for our SSD solutions, but with the open-channel SSD in commercial deployment, our channel SSD will return to growth next year.

Our Ferri SSD should also return to growth.

Now I will turn the call over to Riyadh to discuss our financial results and our outlook.

Thank you, Wallace, and hello, everyone.

I will summarize our financial results and then 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.

Please note that all non-GAAP results exclude FCI in order to provide transparency to the performance of our continuing operations.

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

In Q2, revenue grew 6% sequentially.

Results for our 3 products are as follows.

SSD controller sales grew about 15% sequentially.

eMMC+

UFS controller sales grew about 20%, and SSD solutions declined about 40%.

Gross margins in Q2 were up sequentially at 51.5% compared to 50.2% in the prior quarter as sales of our higher margin controllers grew while sales of our lower margin SSD solutions declined.

Operating expenses in Q2 rose to $31.1 million from $28.6 million in the previous quarter due primarily to higher R&D compensation expenses.

Operating margin in Q2 increased to 18.4% from 18.0% in the prior quarter due to higher sales and gross margins.

Our effective tax rate in Q2 was 3% compared to 14% in the prior quarter, lower primarily due to a one-time tax benefit which came from the reversal of a previous tax accrual.

Earnings per ADS in Q2 increased to $0.52 from $0.42 in the previous quarter.

Stock-based compensation in our operating expenses, which we exclude from our non-GAAP results, was $0.3 million in Q2 compared to $4.1 million in the prior quarter due to seasonal timing of RSU awards.

We had $345 million of cash, cash equivalents and short-term investments at the end of Q2, up $64 million from the prior quarter.

$44 million of the incremental cash was related to payment at the closing of the FCI sale.

Cash flow from operations generated $40.2 million in Q2.

This quarter, we had $2.9 million of CapEx, primarily for R&D test equipment and design tools.

In late May, we paid $11 million or $0.30 per ADS in dividends to shareholders.

This was the third quarterly installment of our $1.20 per ADS annual dividend that was announced in April of last year.

We did not repurchase any shares this quarter.

Now let me turn to our third quarter guidance, outlook for the rest of the year and business trends leading into next year.

In Q3, we expect revenue to grow 10% to 15% sequentially, driven by our 3 key products, SSD controllers, eMMC+

UFS controllers and SSD Solutions.

Current indications from customers suggest our Q4 revenue could grow 15% to 20% sequentially.

Let me provide color on our 3 key products.

We expect our SSD controller sales to grow sequentially in both Q3 and Q4 with stronger sequential growth in Q4.

For the full year, we expect our SSD controller shipments to grow at least as fast as the growth rate forecasted by industry analysts.

Though our SSD controller sales growth will be lower than shipment growth because of product mix shifts, strategic and tactical reasons disused by Wallace earlier.

Our SSD controllers are very well positioned for further growth next year as our OEM programs continue entering volume production and a more favorable NAND pricing environment benefits our module maker customers.

We expect eMMC+

UFS controller sales growth to grow sequentially in both Q3 and Q4.

For the full year, our eMMC+

UFs will decline meaningfully compared to last year and should be flat to up next year as our module maker activities and U.S. flash partners' UFS growth offset downside from eMMC transitioning to UFS at our Korean flash partner.

We expect our SSD Solutions sales growth to grow sequentially in both Q3 and Q4 with new Shannon open-channel SSD sales and growth from our Ferri SSDs.

For the full year, however, our SSD solutions sales will be down sharply as our legacy Shannon direct I/O SSDs decline faster than expected and procurement plans for open-channel SSDs have been reduced.

Also, our Ferri SSD sales declined as NAND prices fell more than expected.

We expect our SSD solutions to grow next year because we are now well positioned to introduce our next generation open-channel SSDs for our 2 B-A-T customers' 2020 projects.

With more favorable NAND pricing conditions next year, we also expect our Ferri SSDs to return to growth.

Q3 gross margin is expected to be between 48% and 50%, lower than the 51.5% in Q2 primarily because of ramping lower gross margin SSD solutions sales.

And secondarily, because of slightly lower SSD controller margins.

We expect Q4 gross margin to decline further because of the two same reasons.

Full year gross margin is expected to be between 47.9% and 49.9%, roughly similar to the 49.1% last year.

Q3 operating margin is expected to be between 17.3% to 20.3%.

We expect both Q3 and Q4 operating expense to be roughly similar to Q2.

For the full year, we expect operating margin to be in the range of 17.6% to 20.6%, lower than the 27% last year.

Full year operating expense should be roughly 10% higher than last year.

Stock-based compensation in Q3 is expected to be in the range of $2.4 million to $3.4 million, and full year stock comp in the range of $14.8 million to $16.8 million, lower than the $20.2 million last year.

Our effective tax rate in Q3 and Q4 is expected to be approximately 15%, our model tax rate.

Before I wrap up, let me brief everyone about impairment issues at Shannon.

We acquired Shannon in mid-2015 to leverage our unique controller technologies to create incremental shareholder value by supplying customized enterprise SSDs.

Analogous to our customized Ferri industry SSDs, and also a vehicle to accelerate our enterprise SSD controller development.

Many aspects of our Shannon operating performance have come in below expectations.

The competitiveness of our legacy custom direct I/O PCIe, 8HCI SSDs declined faster than expected and they are now superseded by standard PCIe and NVMe SSDs.

And our custom open-challenge NVMe SSDs entered production a year behind schedule and at significantly lower volume than planned.

Our overbuild of products and overstock of NAND components combined with rapidly falling NAND prices have led us to write off $5 million of inventory this quarter.

Additionally, a significant portion of our $34 million of goodwill and intangible assets from our Shannon acquisition is now likely impaired due to lower projected future profitability and cash flow.

We are still conducting our impairment evaluation and we expect to take a large impairment charge later this year.

Finally, you should expect us to be repurchasing our shares this quarter.

This concludes our prepared remarks.

We will now open the call to your questions.

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

Riyadh and Wallace, perhaps a more technological question initially.

Your client SSD business has been doing quite well, but as the industry transitions from SATA to NVMe protocols, may you comment on how you view your technological advantage relative to your customers, some of which are keen on building up internal design team controller resources?

And I guess if you truly have a technological advantage, why the need to lower controller prices to staff off competition?

Then I have a follow-up.

Regarding our lowering our controller costs during the period due to NAND price decline sharply and most of our module maker and some of our NAND makers shipping to OEM customers adopt DRAM-less PCIe NVMe solutions.

So our average ASP for this controller is lower than the DRAM based PCIe controller.

Second is, due to a majority of our module maker SSD customers, they cannot make a profit.

So we really want to make sure they can survive and continue working on the SSD business in the industry.

So diversify is our strategy.

We'd like to see they can rebound gradually from second half and be successful in 2020.

That's why we need to reduce our ASP to meet our customer expectation and need and to build a strong infrastructure for SSD bottom line.

Karl, let me also add to what Wallace had said and just sort of summarize.

There's 3 big categories of what's happening with our ASPs and we're really talking about blended ASPs.

The first relates to product mix.

The second is about strategic decisions we've made in order to maintain our diversified ecosystem.

And the third is tactical to drive off the small competitors who are seeking a foothold in our marketplace.

So, as Wallace mentioned, the first is relating to product mix.

SSDs have become so performance oriented relative to hard disc drive, already the performance is so much better than hard disc drive that it's not about just performance.

It's also about cost.

And this is leading customers, initially module makers and now also flash makers, to seek lower cost controller solutions.

In the past it was about module makers using DRAM-less solutions for SATA drives.

Now even the flash makers are using DRAM-less solutions, but DRAM-less solutions for PCIe and NVMe.

So OEMs moving from SATA to PCIe and NVMe, but as they move to NVMe, they are also adopting a lot of DRAM-less solution which is causing our blended ASPs to shift to the lower price points.

So it's not so much about reducing our price, it's about our blended ASPs coming down because of our customers blending to lower cost parts given the already very favorable performance that you can get from the DRAM-less solutions as they have converged from better controller designs that we've introduced giving them this capability.

That's very helpful.

As my follow-up, Riyadh, thanks for the color on the segments.

Is the lower procurement for Shannon SSDs an admission that the design completion missed the production window for your custom SSD products to both B-A-T providers in the second half of this year?

And I guess if it is, how should we think about the commercial deployment opportunity both in the second half of calendar 2019 but also the first half of calendar 2020?

Thank you.

So we are in the real commercial production with the real workload in the B-A-T customers in second half.

So this is about around 6,000 SSD units with 1,000 servers, the real workload running.

I think after so-called [great task] next year, I think you can expect the order will be around 600 petabytes to 1 zebibyte range.

And this is the same -- because this year, we don't want to be too aggressive in Shannon open-channel due to the price decline and certain procurement strategy from our customers.

So we really purpose and make sure it goes through the real workload rounding the real system with SATA server, then we become the major supplier to 2 of the major NAND makers together.

So this is our plan and with that we are looking forward to strong recovery for 2020 next year.

Could you please repeat your second question?

Sure, Riyadh.

How should we think about the commercial deployment opportunity especially in SSDs in the first half -- both the second half of calendar 2019 and versus the first half of 2020.

I guess how should we think about the linearity of product momentum perhaps first half of 2020 versus second half of 2019 on what appears to be a nascent but depressed base in the second half of 2019?

Thank you.

For this year, as Wallace mentioned, our products are now in commercial deployment at our 2 B-A-T customers.

But bear in mind, the lower levels of procurement by our 2 customers this year is a big part of the result is our first-generation product is frankly massively delayed.

Originally, we were planning to bring this into production in the second half of the year, but open-channel SSD design and performance tuning has taken a lot longer.

It's a lot more challenging than we had originally anticipated.

But we have finally overcome all these issues and have brought our products into commercial deployment.

But given that we are already well past the half year mark for this year, the procurement needs from our customers have been dialed down.

So what we're doing now is more gearing ourselves up for next year.

So we're very excited about next year given that we've already gone through a lot of the learning, gone through some very sharp learning curves to get to where we are today.

So we are now getting our second generation of our open channels ready for testing and qualification but they should be a lot smoother this time around.

And gearing ourselves to bring our products, our second-generation products, to our 2 customers.

Your next question comes from the line of Anthony Stoss from Craig-Hallum.

Hi, guys.

Riyadh, can you tell us how much revenue Shannon systems was in Q2 or what percentage of revenues?

And given it's been very lumpy, they haven't executed to plan, what can you do to further cut costs within that division?

Then I have a couple of follow-ups.

Revenue level at Shannon is now down to bare bones.

The products that Shannon used to sell were these legacy direct I/Os, so these are the pre-PCIe NVMe drives with custom driver software.

Those have largely been superseded by PCIe NVMe and so the sale of those products for example at Alibaba ended last year.

We still had some legacy customers that we've been shipping to in Q1 and Q2, so we're down to bare bones.

So we're now gearing up for the sales of our NVMe, our open channel NVMe drives in Q3.

These are already in commercial deployment.

We are already shipping and these will ramp in Q3 and further in Q4.

Where do you need Shannon systems just a group to breakeven?

What kind of quarterly revenue run rate?

Shannon has been breakeven in past years, but this year is not given that revenue levels have fallen so much.

But we are expecting Shannon to return to breakeven next year.

Okay.

Then kind of bigger picture on gross margins, for years you were always targeting 50% to 52%, somewhere in that range.

Granted it's only for the quarter kind of 48% to 50%, is that the new long-term gross margin perspective now that you cut prices and really can't take that back?

Is 48% to 50% the new level going forward?

It depends on product mix.

We have -- the ASPs of our SSD controllers have come down a bit and the margins have also come down a bit, largely from product mix as our customers have shifted to lower cost components in order to sell more attractively priced lower cost SSDs.

But as we move into next year and the following year, we're expecting our enterprise SSDs to start contributing and that's going to have a positive momentum both in terms of our ASPs as well as our margins.

And broadly, given that our products are highly diversified, we are always seeking to bring higher valued products to our customers, means for us to continue to bring more value, and hence the opportunity for us to blend up our gross margins.

One last question if I may, last quarterly conference call and throughout the quarter you talked about turning away business on the SSD, client SSD side because lots of design folks are coming to you and you can't handle it.

How does that rectify versus your guiding level is full year maybe you're only going to grow in line with the market, versus again, taking resources from say Shannon systems and deploying them more towards the client side?

First of all, the Shannon line SSD resources, they are focused on enterprise.

So the scale in knowledge is quite different than on the client side.

We definitely are continuing with our resource and also use our resources smartly, assigned to the major OEM customer projects.

We believe we have been through the worst condition in the Q2 because now most of our product line from client SSD, firmware and all kind of NAND are almost ready.

That's why we see the customer and because NAND price stable, so picking up the demand and forecast, our visibility is much better than the first half.

Now we do see if the NAND pricing remains stable, supply sufficient, we then can see continued growth on both our NAND OEM customers as well as module maker customers.

Your next question comes from the line of Craig Ellis from B. Riley FBR.

Wallace, Riyadh, thanks for taking the questions and good evening.

I wanted to start off just by following up on some of the gross margin questions.

It looks like if we just back into the math in the fourth quarter that gross margins are expected to be in the 46% range.

So one, is that correct?

And two, Wallace, can you just frame the risks that 3 drivers that are impacting gross margins in the near-term could either become more adverse exiting this year and become more of a headwind to a rebound in gross margins as we go through 2020?

Or conversely, work more to your favor and frankly provide upside to what is I think an expectation for another 100 or 200 basis points stepdown in gross margin in the fourth quarter?

From pure controller portion, we believe through the proper product mix in the near future when we issue more PCIe controllers SATA, we issue more high-end PCIe controller than value line PCIe, our gross margin will go higher.

And when we issue more UFS 3.0 than UFS 2.0 controller, our gross margin blended basis will be higher.

So that will be potential contribution for our gross margin growth.

However, when we ship more SSD solution, our Ferri base solution and that product line, these product lines cannot maintain our corporate average gross margin.

And that will drift down our average gross margin.

So it really depends on product mix.

So I see in the future, if our solution, our SSD solution and the Ferri base can grow much faster, I think we will look at the dollar as a really profitable result.

Not just gross margin.

But from controller side, we believe in the near future, especially 2020, we will see more PCIe controllers and UFS 2.1 move to 3.0, that will benefit for higher ASP and better gross margin.

So controller portion sales revenue, we should have a better gross margin than 2019.

Craig, let me also go back to your first question, your earlier question about our fourth quarter gross margins.

The number you quoted is roughly in the ballpark of what we're expecting in Q4.

Normally when we look at our gross margin, it's really the mix between our solutions and our controllers.

Our controller gross margins are generally very stable and above corporate average, while our solutions are generally meaningfully below our corporate average.

But this time around, in the second half of the year, our SSD solutions gross margins are going to be extremely low, especially because of our Shannon products.

As you may recall from our prepared remarks, we've just written off $5 million of Shannon inventory which means we're essentially going to be selling the rest of the inventory which are now marked to fair market value at close to zero gross profitability.

It's still positive, but it's just a few percentage points above zero.

So as these products sell and ramp in Q3 and further in Q4, it's going to drag down our overall blended gross margins.

This issue will go away as we consume all of these expensive NAND components and SSDs that we built in past quarters.

Once we consume all these and we bring in newly purchased NAND components that more reflect today's NAND prices, our gross margins for our SSD solutions will go up meaningfully.

But even when they go up meaningfully, they are still going to be below our corporate average levels.

Thanks for all the detail there, that's helpful.

The second question is regarding some of the comments that were made regarding the business with flash OEM customers which had a very nice performance in the quarter.

One of the comments that was made was that you see continued signs of outsourcing interest with those customers.

Can you provide further color on whether that's a continuation of trends that you've seen in the past or if there is something new that's happening that provides you with visibility into new vectors of growth, either as you look at the back half of this year or next year?

At the moment we cannot comment the details, but we have very strong engagement with some other NAND makers.

We do see they plan to outsource to third parties in 2020.

So we started dialogue engagement and hopefully they reflect to our 2020 financial performance.

Got it.

And then lastly, nice to see that the company will be taking advantage of current stock prices with what is a very robust cash position.

Can you give us some sense for the intensity with which you intend to approach the buyback or if the board has agreed on a certain buyback amount that the company would target either this quarter or next quarter, how you would approach the use of cash?

Thanks, guys.

We have a $200 million share buyback program that was authorized in the fourth quarter of last year.

It's a 2-year program.

We have already used roughly $35 million of the $200 million authorization.

Every quarter before we put our buyback program into deployment, we have to get cash disbursement authorization from our board and as part of that authorization, our board will discuss with us about our business visibility and the terms for the upcoming quarters' purchases.

So we generally take a measured approach.

As you know we are generally fairly conservative of how we deploy our cash.

But at the right moments, we will buy our shares and you should expect us to be in the market this quarter repurchasing our shares.

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

Thanks for taking my questions.

Going back to the client SSD business, just trying to reconcile the commentary.

It appears that we're reaching kind of a saturation point with respect to performance of SSD versus hard disc drive and therefore the OEMs or the customers are not willing to pay up a premium over hard disc drives anymore because performance now is already intact.

How do we think about that going forward with respect to overall ASPs?

I understand you're hoping to ship more PCIe and other higher margin products within the client SSDs, but it appears to be that there is going to be kind of a structural change in the margin profile of that business going forward.

So I kind of wanted to get a better understanding, is this more a structural change now in the industry?

A change in the structural pricing?

Or is there kind of a hope that this can be reversed through a better mix of products?

I think that's a very good question.

We always in also planning for this direction.

As you know, the market is changing.

PCIe Gen 4 is coming.

Although AMD are launching Q1, but Intel products will launch around late Q3/Q4 next year.

So the PCIe Gen 4 will bring a new category or product line to compete in the market.

However, for PCIe OEM in commercial line, the (inaudible) will stay with 512 gigabyte to (inaudible) gigabyte because they are for opposite users.

For consumer product line, I think you are going to see 1 terabyte and 2 terabyte and they are going to move to higher density to meet a consumer expectation.

So the NAND will diversify, one is performance driven, one is cost driven.

And our controller, we see we can add more value in QLC based SSD than controller itself, the cost structure will be higher.

Because we need to increase performance to meet a PCIe Gen 4 category.

In the same token, if based on TLC base, we can really leverage our technology to create more creative way moving from mainstream product like a premium line product, then our controller has more value and customers are willing to pay for the ASP.

So from blended basis, we need to carefully leverage our (inaudible) SSD controller product line so it can give us overall the price contribution and margin contribution can be higher.

And we see that's a trend because NAND is going to (inaudible) layers more to 120 layers next year and we will benefit more because our controller will be ready ahead of most of our competitors in this market.

Raji, let me also add to what Wallace has said.

Just to summarize, there are 2 technology trends that are taking place and one of the trends is just to bring low cost SSDs to displace hard disc drives.

So this is more of a commodity play, especially when you have the NAND flash makers with a lot of NAND that they need to dispose of and consume.

The other trend is all about performance.

You have the PC OEMs already using SSDs for their more performance-oriented devices.

And for that part of the market, it's about performance, bringing higher and better performing SSDs to them.

So right now, the balance is more towards the cost side, but over time we expect it to shift back to performance when the balance moves in the other direction.

Thanks for the insight.

Just to follow-up, roughly what percentage of your client SSD revenue is to module makers versus OEM, just to get a sense of the impact of that business because of the falling NAND pricing?

I guess along those lines, I wanted to get a sense of what your view is now of the attach rates, the overall attach rates of client SSDs in the PC market?

Any sense in terms of what the percentage of that is now?

I think for this year, particularly this year, OEM sale revenue for clients be higher than we sell to module makers, so that's probably around 60% is for OEM customers.

But normally in the past, we are roughly 50/50 balance between PC OEM and module maker customers.

Regarding the [attach rates], I think around about 55% to 60% of all PCs, including desktops, adopted SSD today.

So they are still around 50%, 40% of historic device is opportunity for SSD to replace HDD in the next few years.

Raji, let me also add, the market is not just PCs.

While the PC market, we've seen the highest adoption of SSDs, we also have consumer electronics and external storage.

Consumer electronics include game consoles like the Xbox and PlayStation.

The adoption of SSDs for those devices have not even started.

And then you also have the external storage and surveillance markets.

And so we're looking at 220 million to 250 million client HDDs that are still shipping today for PCs, consumer electronics and external storage.

And all these are opportunities for conversion to SSDs.

Your next question comes from the line of Mehdi Hosseini from SFG.

Two questions.

My first one regards (inaudible).

You have had about $10 of net cash per share on the balance sheet for quite some time.

It seems to me based on what you have reported on a pro forma basis in first half of 2019 and outlook, it seems to me that the peak earning happened in 2016.

I understand the challenges out there in the market, but I want to better understand your strategy how you are going to utilize your cash other than just the capital return.

Is there anything that you are contemplating or are there different scenarios in your long-term strategy that could be enhanced by committing more cash and helping with better earning prospect?

Then I have a follow-up.

Mehdi, generally the cash on hand is intended to provide adequate working capital for our business, support continuing dividend programs and opportunistic share repurchase.

We also look at having the cash on hand for opportunistic acquisitions.

It also provides, the cash on hand also provides peace of mind to our large OEM customers, all of whom are much bigger than we are.

They also provide peace of mind to our foundry partners and employees that we're going to be around for the long haul.

Especially since we're in a business that is very dynamic and volatile.

So for the two key pieces of cash beyond the cash related to just running our business, our dividend, share repurchase, the dividend part of it, we're going to revisit again at the end of the year.

Our current program, we've paid our third installment, so one more installment and at the end of the year our board will then have to decide whether to keep the same dividend amount or to raise it.

Another is, opportunistically we will continue to repurchase our shares, so that is another use of cash.

And thirdly, we do look at M&A opportunities, but those are hard to predict, so it's also important to keep our powder dry.

Sure, so business as usual.

This follow-up is for Wallace.

I see some disruption in the NAND supply somewhere between 20% to 30% of the available NAND next year would have to go through a change in architecture from one technology to other floating gate to charge trap.

Is that also creating some headwind in forecasting for you?

Is there another reason for you to be rather conservative in your embedded solution?

And I'm just wondering, how do you see those changes impacting your 2020 opportunities?

I cannot comment for specific NAND makers regarding technology change or their strategy.

But what I think I can comment is I believe the transition could be very smooth.

We don't think there will be any obstacle challenging for our business to grow for 2020.

We believe we work with all NAND makers including China NAND maker and we plan to have a good rebound in 2020.

In other words, if there is an unanticipated disruption as we go through this technology change, that could cause another source of challenge or overhang for your embedded solution?

I don't get a sense that that's being done.

I think the NAND maker itself, they always have a today calling production and the next generation preparing for future production.

So they will make a judgment to allocate the capital investment, also the capacity output, the NAND output, based on their technology maturity.

They will not, they cannot change generation just in one night.

So they were based on transition regularly for the new technology and moving based on the yield, based on the cost structure and transitioning from current generation to next generation.

So I believe transition will be smooth and based on whenever they are ready, they will, based on their cost structure, moving the capital transition from current floating gate to charge trap for the next generation

Mehdi, let me also add, we see a lot of technology transitions in our industry over the course of the last decade or so.

We've seen transition from 8-inch wafers to 12-inch wafers.

We've seen the transition for SLC to 2 bits MOC to TLC and now to QOC.

We've seen a transition from planer NAND to 3D, we've seen the transition from floating gate to charge trapping.

And generally, they have been, the 6 and 7 NAND flash makers have made the transitions very smoothly.

The industry is full of profit maximizing business leaders with great engineers, and they've generally done a phenomenal job in transitioning technology, one technology after another.

And we also see just the more routine in the past about geometry strength and now it's adding more and more layers.

They've all continued to be very smooth.

And as a controller supplier, we are very well positioned to help them sort through the issues for them to bring their components to the marketplace whether for their own products or to sell their NAND to module makers.

I would actually argue that the past technology transitions were net positive for Silicon Motion as it had some positive impact on changing the controller.

But the upcoming transition is more of a device and I'm not sure if it goes as smooth.

At best case may not have a positive impact.

But if there is any interruption, could limit a bit the availability.

So Mehdi, first, we cannot comment on NAND maker what is the strategy they plan to do.

How they allocate, what portion is folding gate, what portion will be charge trap.

But what I can tell you is, we are working with our partner, so for both mobile and client SSD we have same controller firmware but different generation NAND simultaneously.

So that's why we are sure transitioning will create an impact for the business.

We are not creating that bar business.

Your next question comes from the line of Sujeeva Desilva from Roth Capital.

Wallace and Riyadh, I think if I'm correct, you guided the SSD controller to actually grow faster in the fourth quarter than the third quarter.

Can you talk about the factors you're seeing that are driving that?

Is that seasonality, the lower NAND pricing driving elasticity, or just secular since the attach rate?

Any color there on the visibility into the fourth quarter would be helpful.

I think entering the second half, we have much better visibility.

Our OEM customer, NAND partner have better visibility.

So it gave us forecast in Q4 is looking much stronger growth than Q3.

That's why we are confident Q4 have better outlook than Q3.

Plus, Suji, we have our programs one after another lining up to go into production with our OEMs.

So we have visibility related to that.

And finally, with NAND prices beginning to show signs of stabilizing, we're also seeing greater confidence from our module makers to want to be more active in building SSDs.

So all these are adding to improving expectations for the rest of the year.

Okay.

Then a question on the cost, on the gross margin guidance.

I'm wondering if this is a lever you can pull, but is there any ability to cost down your products to help offset the gross margin pressures in the near term or is kind of the cost structure of your products locked in relative to your kind of product roadmap?

What we do, we are working hard in all angles to reduce our manufacturing costs.

But in the same time, I think our channel on channel solution is going to ramp up in Q4 with more meaningful volume.

That's why we diluted our gross margin, average gross margins.

But we believe as we have said, up until we use all existing NAND and for next year we believe our gross margin per channel solution will be much better.

So the blended basis which you have better solid gross margin moving to 2020.

Last quick question on SSD solutions, is there any opportunity for customers beyond the two China B-A-T customers you have or should we think of those 2 being the drivers of 2020 revenue?

We do have customers beyond the 2 B-A-T customers in hand.

Any color?

Is that a U.S. hyperscaler or any color there?

They are both in China and U.S.

Your next question comes from the line of Charlie Chan from Morgan Stanley.

Wallace and Riyadh, my first question is about the eMMC PCs outlook.

Because Hynix continues to be cutting 2D NAND production still, so any chance you can get into their 3D NAND eMMC or UFS in the coming quarters?

I think of course our Korean partner, their transition to 3D NAND is more related to capacity.

We are still working with our Korean partner and they are also looking for the high-end UFS product because they need to evaluate which controller is better, another third party's controller or ours.

But they have a different strategy.

We maintain very close relationship with our Korean partner, and not just in mobile business, but also in the SSD business.

And the enterprise side too.

So I think I cannot comment and say I think we will grow faster in the Korean partner customer, but it may not be the current Korean partner, it could be more Korean partner customer in 2020.

Okay.

And then client SSD ASP, I mean I can understand the margin pressure, but in terms of blended ASP given you are still growing PCIe mix and that still should be at a big premium to SATA still.

So in terms of blended ASP, can you give some range of the absolute dollars or the trend in 2 coming quarters?

Charlie, we don't think it's that meaningful to talk about our ASPs for our parts.

They go through a wide range.

They have their own product lifecycles.

But what we can comment about is the blended ASP trends of our business.

This year, as our products transition from SATA to PCIe, they're also transitioning, a lot of our growth is now for the PCIe parts of the DRAM-less variety of the PCIe versus the with DRAM.

So we're talking about SSDs using PCIe for more mainstream and lower cost entry level products.

And so the ASPs for those are going to be lower.

Furthermore, as we mentioned in the past, we've provided more favorable pricing of our SATA SSD controllers to our module makers.

And so these have caused our overall blended ASPs to be lower than what we had expected at the start of the year.

Okay, thanks.

And lastly, Ferri PCIe, I thought it should be less sensitive to the NAND price, right?

Because when NAND price comes down, consumer product may be half of it (inaudible) and when NAND price goes up, it will be negatively impact on demand.

But for Ferri I thought the end market should be more like industrial.

Why that is softer than NAND price decline?

Charlie, you are absolutely correct.

Our Ferri was in the industrial sector.

However, entering 2019, due to the NAND inventory and price decline, we do see NAND maker gradually enter into the market with unbelievable price.

That's what impacted our price position.

However, we do, because we do start to see the rebound and growing from the server side as well as automotive sector, and we should see quite strong rebound for our Ferri business for 2020.

And when NAND price going to the more stable condition.

Charlie, let me also add, since NAND prices, don't forget that the price of NAND have come down so much compared to last year, when if our unit shipments of our Ferri products were flat or up and our densities remained fairly stable, the amount of revenue that we're going to be receiving is going to be a lot less than what we were able to enjoy last year.

And so these are factors that you have to consider why you look at our business.

There are no further questions at this time.

I would now like to hand the conference over back to Wallace.

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

We will be attending several investor conferences in Asia and US during this quarter.

Details of these events will be available on our website.

Thank you for joining us tonight.

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

Thank you for participating.

You may all disconnect.