威騰電子 (WDC) 2020 Q1 法說會逐字稿

Ladies and gentlemen, thank you for standing by, and welcome to Western Digital's First Quarter of Fiscal 2020 Conference Call.

(Operator Instructions)

I would now like to hand the conference over to your speaker, Mr. Peter Andrew.

Please go ahead, sir.

Okay.

Thank you, and good afternoon, everyone.

Before we begin, let me remind everyone that today's discussion contains forward-looking statements including product development expectations, business plans, trends in financial outlook based upon management's current assumptions and expectations, and as such, does include risks and uncertainties.

We assume no obligation to update these statements.

Please refer to our most recent financial report on Form 10-Q filed with the SEC for more information on the risks and uncertainties that could cause actual results to differ materially.

We will also make references to non-GAAP financial measures today.

Reconciliations between the non-GAAP and comparable GAAP financial measures are included in the press release and other materials that are being posted in the Investor Relations section of our website.

With that, I'll now turn the call over to Steve Milligan, our CEO.

Thank you, Peter, and good afternoon, everyone.

Joining me today are Mike Cordano, President and Chief Operating Officer; and Bob Eulau, Chief Financial Officer.

Before we discuss our results for the first quarter of fiscal 2020, I want to spend a moment talking about the other news we announced today.

After a long and fulfilling career with Western Digital spanning 2 decades, I have informed our board that I plan to retire as CEO.

I will continue to serve as CEO until the board has identified and appointed a successor.

Western Digital is a significantly different company than the one I first joined in 2002.

We are more diversified, more resilient and much better positioned to capture the opportunities of today's evolving data marketplace.

Since my appointment as CEO, Western Digital has transformed from a storage component provider to a diversified enabler of data infrastructure with the broadest portfolio in the industry, offering customers a powerful combination of hard drive storage and flash memory products.

We have successfully executed many key strategic initiatives including the company's acquisition of SanDisk, the integration of Western Digital, HGST and SanDisk as well as the extension of Western Digital's 19-year partnership with Kioxia.

We now operate a powerful platform that uniquely positions us to provide new architectures and capabilities to manage the volume, velocity and variety of data.

As we think about what comes next for our company, I believe now is the right time for Western Digital to begin the transition to its next phase of leadership.

Serving as Western Digital's CEO for the past 7 years has truly been one of the highlights of my career.

I want to thank our team for their support and dedication.

I could not be prouder of what we have accomplished together and consider myself quite fortunate to have worked alongside such a talented team.

In terms of next steps, I look forward to continuing to work closely with the team while the board conducts the search for our next CEO.

Given our strong management team, we expect this transition to be seamless for our shareholders, our employees and the customers that rely on our best-in-class service and products.

Once my successor is on board, I will remain with the company in an advisory role until September 2020 to ensure a smooth transition.

I will also continue to serve as a director on the Western Digital Board for a transition period after my successor is appointed.

With that said, let me turn to our performance for this quarter.

Fiscal 2020 is off to a good start.

Revenue exceeded the guidance range we provided in July and non-GAAP EPS was at the upper end of the range.

The upside was driven primarily by the success of our capacity enterprise drives for the data center.

We are executing well in the data center, utilizing the power of our portfolio.

At the core of our success in this market is our industry-leading capacity enterprise hard drive solutions and the exceptional value we provide to our data center customers for their diverse storage needs.

During the quarter, we made 2 important announcements to further extend our product leadership.

First, we introduced 16- and 18-terabyte CMR drives and a 20-terabyte SMR drive, all enabled by our energy-assisted recording technology.

These drives are expected to sample this quarter.

Additionally, through our continued investments in heads, media and mechanical design, we began shipping an air-based 10-terabyte drive providing significant benefits to our customers.

I am pleased to announce we commenced the initial revenue ramp of our NVMe-based enterprise SSDs to major hyperscale and OEM customers during the September quarter.

Efforts to qualify and ramp additional customers with our next-generation products based on our 96-layer, 3D flash technology are going well and should position us to further increase our participation in this important market.

Western Digital's ability to offer both hard drive and flash-based solutions differentiates us from our competitors and allows us to more strategically partner with our data center customers.

Outside the data center, the overall demand environment across the consumer, mobile, PC and retail end-markets is solid.

Furthermore, we are seeing improving trends across our flash product portfolio and continue to believe that the flash industry has passed a cyclical trough.

With a broad and growing product portfolio, Western Digital remains well positioned to benefit from the long-term drivers of the growth and value of data.

With that, I will now ask Mike to share our business highlights.

Thank you, Steve, and good afternoon.

Before I get into my prepared remarks, I want to congratulate Steve on his upcoming retirement.

I value and appreciate the partnership we have built together over the past decade and want to acknowledge Steve for his leadership and numerous contributions to Western Digital.

We have quite a bit of time and work to do between now and September of next year, and I look forward to working together to execute on our plan.

As Steve mentioned, fiscal 2020 is off to a good start.

We had record hard drive exabyte shipments driven by the success of our capacity enterprise drive family.

We also had record exabyte shipments in flash, as we benefited from demand elasticity and share gains in SSDs for PCs and notebooks.

In Data Center Devices and Solutions, our capacity enterprise exabyte shipment growth was over 60% year-over-year led by the ramp of our 14-terabyte drives.

These drives now represent the majority of our capacity enterprise unit and exabyte shipments.

Industry analysts expect the 14-terabyte capacity point to be the industry's highest volume product through the first half of calendar year 2020.

Building on our areal density leadership and execution on mechanical design, we announced our plan to accelerate the introduction of our and 9-platter, energy-assisted capacity enterprise drive platform.

This enables us to ship 16- and 18-terabyte CMR and 20-terabyte SMR drives on a unified platform, simplifying the qualification process and reducing the time to market for our customers.

We will be sampling all of these drives by the end of this quarter and will commence volume shipments in the first half of calendar 2020.

In addition, we began shipment -- shipping a new 10-terabyte air-based product powered by our innovative airflow architecture, underscoring our areal density and mechanical design leadership.

Given the strength of our capacity enterprise portfolio and the opportunities we see in this market, we now believe our exabyte shipment growth will exceed 40% in calendar year 2019, up from our prior estimate of 30%.

In enterprise SSDs, our NVMe-based products experienced a strong quarter of growth, and we expect continued growth in the December quarter.

We are qualifying our next-generation 96-layer product with additional customers, which positions us for further market share gains in calendar year 2020 and beyond.

We have a unique and sustainable competitive advantage within the data center built on strong customer relationships and a strong product portfolio.

Our strategic position within this important end market will drive future revenue growth.

In Client Solutions, revenue grew on a sequential basis driven by an improving pricing environment and a seasonal increase in bit shipments.

This quarter, we began shipping 96-layer QLC-based retail products and external SSDs.

The trust and reputation of our brand and our customer's preference for the performance and reliability of our solutions are key differentiators.

In Client Devices, the main contributor to our year-over-year decline was our decision to limit our participation in the mobile market.

On a sequential basis, we expect to ship more bits into the mobile market in the December quarter.

In PCs and notebooks, we gained market share in client SSDs as our exabyte shipment growth exceeded 70% year-over-year.

Our bit production of 96-layer BiCS4 surpassed 64-layer of BiCS3 during the September quarter.

We are on track to commercialize BiCS4 across all our product lines by the end of calendar 2019.

Our JV partner, Kioxia, and Western Digital executed well, bringing the [old pet fabs] back to full production after the power outage, limiting our output reduction to 4 exabytes.

Our flash supply is tight, and we continue to believe that any excess inventory in the flash industry supply chain will be substantially reduced by the end of calendar 2019.

We expect flash industry supply bit growth to be in the mid-20% range in calendar 2019 and in the low 30% range in calendar 2020.

In addition to continued growth in our existing flash portfolio and capacity enterprise markets, we see several new incremental growth opportunities.

First, the launch of the next-generation gaming consoles will be important events for the gaming industry and our flash business.

We expect these new consoles to utilize high-capacity flash storage to improve the gaming experience.

Second, we expect to expand our product portfolio and diversify our customer base for the mobile market with our UFS, eMMC and custom solutions.

Finally, our recent announcements of 3D flash products for the automotive and industrial markets will further expand our opportunities in these growing and more stable flash-based markets.

I will now turn the call over to Bob for details on our financial performance.

Thanks, Mike.

I also want to congratulate Steve on his upcoming retirement and look forward to helping enable a seamless transition to the next CEO.

I'm pleased to announce that revenue for the September quarter exceeded the high end of the guidance range we provided in July and non-GAAP earnings per share came in at the high end of the range.

Revenue for the September quarter was $4 billion.

This was up 11% sequentially as we experienced growth in Data Center Devices and Solutions and Client Solutions.

Revenue was down 20% year-over-year as we faced a tough compare -- tough comparable quarter in fiscal 2019.

By end markets, Data Center Devices and Solutions revenue increased 20% sequentially and 6% year-over-year due to the success of our capacity enterprise drives for the data center.

Please recall that a year ago, based on discussions with our customers, we predicted that in the second half of calendar 2019, we would return to growth in capacity enterprise.

We are experiencing that predicted rebound now.

On a sequential basis, Client Solutions revenue grew 18% on seasonally stronger flash bit shipments and a stronger flash pricing environment.

Client Solutions revenue declined 4% year-over-year, primarily due to a reduction in hard drive TAM.

Client Devices revenue was up 1% on a sequential basis and decreased 39% year-over-year.

The year-over-year decline was a result of our decision to scale back flash bit shipments to the mobile market, flash price declines and a reduction in the hard drive TAM.

By product category, flash revenue was $1.6 billion, up 8% sequentially and down 36% year-over-year.

Flash ASPs were flat and bit shipments were up 9% sequentially.

Hard drive revenue was $2.4 billion, up 13% sequentially and down 3% year-over-year.

Average price per hard drive was $81.

Exabyte shipments were up 23% sequentially, hitting a new record level.

As we move on to cost and expenses, please note all my comments will be related to non-GAAP results, unless stated otherwise.

Gross margin for the September quarter was 24.8% with a flash gross margin of 19.3% and a hard drive gross margin of 28.5%.

We completed all of our cost of revenue reduction activities we outlined in January, resulting in a decrease of more than $100 million in quarterly spending.

The hard drive gross margin was up slightly from the June quarter, and we expect this December quarter gross margin to be approximately 30% as we fully realize the benefits of the cost reduction efforts.

Flash gross margin was up on a sequential basis as we benefited from a better flash pricing environment.

K1 fab cost was $64 million, higher than expected.

Excluded from the cost of revenue was a $68 million charge related to the power outage.

Operating expenses were $767 million.

Adjusting for a normal 13-week quarter, operating expenses were below the $740 million run rate target.

During the quarter, we completed all of our operating expense reduction efforts announced in January.

In addition, once we complete the exit of our storage systems business, we should start to see an approximately $25 million per quarter reduction in operating expenses beginning in the March quarter.

Operating cash flow for the September quarter was $253 million and free cash flow was $294 million.

Capital expenditures, which include the purchase of property, plant and equipment and activity related to flash ventures on our cash flow statement were an inflow of $41 million.

As previously noted, we are benefiting from the timing of the funds flowing back and forth between us and the joint venture.

For the full fiscal year, we continue to expect capital expenditures that will have flowed through our cash flow statement to be under $500 million.

Total capital expenditures, which includes our portion of joint venture leasing and self-operating funding is expected to be similar to last fiscal year, between $2.5 billion and $3 billion.

In the September quarter, we distributed $147 million in dividends to our shareholders.

We paid down debt by $319 million, which included an optional $250 million debt pay-down.

As our cash generation continues to improve, our first priority will be to reinvest in the business to maximize long-term shareholder value.

After paying our dividends, our next priority will be to reduce our debt.

At the end of the quarter, we have $3.2 billion in cash and cash equivalents, our $2.25 billion revolver remains unused and our gross debt outstanding was $10.4 billion.

Total inventory dollars were flat on a sequential basis but higher than projected as the joint venture fab recovered faster than expected.

This resulted in a sequential increase in flash inventory, particularly at the end of the quarter, while hard drive inventory declined about $100 million sequentially.

Moving on, our non

(technical difficulty)

Ladies and gentlemen, please stand by.

Sherry, can you hear us on your side?

Yes.

Are we okay?

Do you know where we got cut off?

(technical difficulty)

and with guidance.

So again, this is non-GAAP guidance and it's as follows.

We expect revenue to be in the range of $4.1 billion to $4.3 billion.

We expect gross margin to be approximately 25% to 26%.

Please note that this range includes approximately $75 million in costs associated with the K1 fab.

Operating expenses are expected to be between $750 million and $770 million, above the $740 million run rate target due to higher variable compensation spending.

We expect interest and other expense of $85 million.

We -- and we expect the tax rate to be 26% plus or minus 2 points.

As a result of this detailed guidance, we expect earnings per share between $0.45 and $0.65, assuming approximately 302 million in fully diluted shares.

With that, I will now turn the call over to the operator to begin the Q&A session.

Operator, we'll now take our first call -- first question.

(Operator Instructions) Our first question comes from Aaron Rakers with Wells Fargo.

Steve, congrats on the retirement.

It's been great working with you.

Two questions, if I can, real quick.

So first of all, I guess one of the things that stands out a little bit is the capacity shipment number on flash, up only about 9% sequentially.

By my math, it's up maybe high single digits on a year-over-year basis.

So can you talk a little bit about -- it seems like the product portfolio is in a great position, NVMe is ramping, client SSD capacity shipments were strong.

I'm just curious of why we -- it seems to be a little bit muted as far as the capacity shipment trends in flash.

And what's your expectation going into the December quarter for capacity share?

Yes.

So Aaron, let me address that.

So the primary reason for that is really the point that I made and Bob made in our comments is we did not participate in a substantial way in the mobile marketplace in the quarter just completed.

So that's the primary driver of bit shipments in the quarter.

Okay.

And remember -- and recall -- Aaron, recall -- this is Steve, recall that, that lack of participation in the mobile market was by choice from our standpoint, given that the profitability levels for that segment of the market were not at all attractive.

Okay.

Fair enough.

And then as a second question on nearline capacity enterprise drives.

In the slide deck, you note that you now expect, overall, the market to "approach" 30% year-over-year growth.

I think last quarter you talked about growth meaningfully exceeding that 30% level.

I know you guys are talking about north of 40% growth in your capacity shipments, but what's changed over the last quarter?

Has there been a bit of a softening in terms of your expectation, I guess, going into the December quarter?

Or what's really driving that change of growth expectation?

So Aaron, I think for us, we actually updated our performance on the year.

We originally stated last quarter, we'll be north of 30% for us.

We updated that guidance to north of 40%.

We're actually seeing strength in exabyte consumption across the capacity enterprise segment.

Now the other thing that's happening for us that's more unique is the power of the 14-terabyte product is doing quite well.

And obviously, we are gaining market share in that segment that's showing up on an exabyte basis.

Is -- are there constraints in the market, which is tempering the overall market, the industry expectation?

No.

I think we would suggest the industry will grow at north of 30%, we will grow at north of 40%.

And again, that's all up from our last outlook on both numbers.

Our next question will come from Medhi Hosseini with SFG.

Steve, good luck with your next endeavor.

It was very nice working with you.

Moving on to questions.

Can -- just to follow up.

It's very helpful when you talk about exabyte shipment guidance, especially for the nearline.

And as you look into the next year, how do you see that exabyte shipment target changing?

And again, this is for 2020 versus 2019.

And how would -- how should we think about the mix of the nearline exabyte as a percentage of the overall exabyte shipment for Western Dig?

Yes.

So let me just comment specifically at capacity enterprise.

We would expect for 2020, our current outlook is about 35% year-over-year growth.

So continued strength year-over-year.

And so we don't split out total HDD exabyte growth, we don't specify that.

Okay.

And oh, by the way, Mehdi, that 35% is consistent with our longer-term expectation for us.

Sure.

You kind of preempted my prepared question as -- by saying that you didn't participate in the mobile segment as it relates to your NAND shipment, and there's a debate as to what happens to that inventory in the channel reserved to mobile segment as you look into the March quarter.

So with that as a background, how do you see the supply and demand in NAND looking into the March quarter?

And I'm not asking for a guide, I just want to better understand your view.

You didn't participate in the market, and in that context, how do you see your prices trending into March quarter?

Yes.

So I'll take that, Mehdi.

We -- first off, as we indicated, we believe that we've passed the trough in terms of the flash cycle.

And the overall inventory situation is improving.

In other words, supplies getting much more aligned to demand.

And we would expect that largely for ourselves and for the industry as we exit the December quarter, that things will be fairly in balance.

Now when you move into the March quarter, then let me actually broaden that question a bit.

When you move into the first half of the year, one of the things that we have to keep in mind is that we will see a typical seasonal decline in terms of -- from a demand perspective.

Supply is relatively linear.

And so we will have to -- just like we do every -- largely every year in terms of the calendar cyclicality, we'll have to manage through that.

But then as we move to the back half of the year where we will see -- and this is the back half of calendar year '20, we will see that begin to flip, demand begin to improve.

And we'll see -- rather than modest improvement in our financial results that we've been seeing, we should see an accelerating improvement in our performance from a financial perspective as we move into the back half of calendar 2020.

Yes.

Mehdi, I'll just give you some numbers to work with there.

We talk about low-30s on supply bit growth, we would expect demand for the calendar year to be slightly above that.

Our next question comes from Karl Ackerman with Cowen.

Steve, again, congrats on your retirement and best of luck in your future endeavors.

Two questions, if I may.

Thinking of mobile for a moment.

You referenced that mobile margins have not been attractive for the last 2 quarters, but is that because you don't have captive DRAM?

I guess how important is that -- is it for you to have, either captive DRAM or a new long-term supply agreement for DRAM as you contemplate your competitive position in smartphones over time?

Yes.

Let me answer that.

So let me delineate mobile.

So there's discrete and component part of mobile and then there's the MCP, which includes DRAM.

I think strategically, we do not see MCP as a long-term strategic place for us to operate.

We're focusing our

(technical difficulty)

Ladies and gentlemen, please stand by.

The line is working.

Speaker?

Yes, Sherry, can you hear us now?

Yes, we can.

Where were we last time?

Do you want to back up?

Okay.

Let me back up and repeat that.

I don't know where we dropped off.

Our mobile market participation, let me break it into 2 components.

One is the MCP business that includes DRAM, the second is discrete and component participation in mobile end use applications.

Strategically, we've moved away from product investment in MCP and over the longer horizon, it's not an area of product focus for us.

So when we talk about participation, it's that, plus the discrete business which we chose for economic reasons to minimize our participation as we had higher-value places to put our bits.

So we see that sequentially improving and the economics in mobile improving along with other segments of the market.

That's helpful, Mike.

As my follow-up, shifting gears to gross margins for a moment.

Clearly, you and your peers are operating well below normalized run rates in NAND.

At the same time, though, I think your outlook for hard drives -- hard drive gross margins are good, but still a little bit below where we were roughly a year ago.

I guess will the exit of the systems business or areas of the systems business be the primary driver for gross margin improvement in the hard drives?

And I guess how do we think about the margin implications from the incremental disk and heads on those higher-capacity drives?

Yes.

So let me -- I'll address that and then Bob and Mike can chime in and add any additional color.

So the first thing is the exiting in the system business will have no material impact one way or the other on our gross margins.

So if you look at our hard drive gross margins -- I mean let's be clear about that.

Our hard drive gross margins, although good levels, are not where we want them to be.

We want those hard drive gross margins to be north of 30%, in the low 30% range.

We are still dealing with some of the cost overhang of exiting our Kuala Lumpur manufacturing facility.

That is now behind us, and so we should see our hard drive margins improve into that low 30% range as we exit this -- as we exit the December quarter.

And then obviously, our intent is to sustain and possibly improve that over a period of time.

Flash gross margins are clearly not where we want them to be.

They are improving, albeit at a slow rate.

We would continue to expect as we see this sort of ripple through the market because different customers started at different levels, different markets started at different levels.

The pace of that improvement is not linear for all of those aspects, but we'll continue to see steady improvement in our flash margins this quarter and then into subsequent quarters.

And as I indicated earlier, we expect that improvement to improve at an improved rate as -- at a better rate as we move into the second half of calendar 2020.

Our next question comes from Mark Delaney with Goldman Sachs.

I have some follow-ups around gross margins.

And maybe first, just to better understand the outlook for NAND and how to get the guiding for margins to improve somewhat next quarter.

[Please help us] understand the -- around ASPs on a like-for-like basis, if you can give more color on what you're assuming there, given the comments about a cyclical bottom.

Yes, I can start.

And first of all, I want to remind you on the NAND side, we do have a headwind with the K1 startup costs and bringing up that fab.

And we're now beginning production there.

But -- and there's probably in the neighborhood of a 3-point headwind that we're faced with on the NAND side.

And as we've ramped volume, then obviously, the margins will improve, everything else being equal.

So in terms of flash overall, as we get to equilibrium between supply and demand, we're definitely expecting that the pricing will get better as we move through '20, like Steve was saying.

So I think it will take a little bit of time to work through that, but I think we're going to be in a good place.

Okay.

And then my follow-up was on the hard drive gross margin, again, along the lines of a -- the prior questioning.

I have been under the impression that for the December quarter, hard drive gross margins could hit 30% especially given the upside that the company's seen in the nearline business, which I think typically runs at least 30% if not higher.

Is there anything in terms of increased headwinds around gross margins that the company has seen in the December quarter that's maybe keeping hard drives' gross margins under 30%?

Or was I of the wrong impression about the ability to hit 30% with my previous expectations for the December...

No, let me -- I'm sorry.

Let me -- because I get kind of -- this one I feel strongly about.

We will -- we are intent as to have gross margins north of 30% for our hard drive business this quarter, the December quarter.

And so there's no headwind at present, of course, there can be things that will happen, but there's nothing at present that indicates that we won't hit that level.

And the other thing I would add, Mark, is if you look out over time, we expect more and more growth in terms of capacity enterprise.

Capacity enterprise will become a bigger percentage of our overall mix, and that will help the margins go up as well.

Got it.

That's it.

And Steve, good luck with your future endeavors.

Thank you.

Our next question comes from Mitch Steves with RBC Capital Markets.

I think I want to focus a bit just on the gross margin, kind of the NAND inflection.

I think a lot of people are looking for kind of like $0.80 to even $1 for the December quarter guide due to improving memory environment.

So I guess maybe, can you help us at least understand how you guys think of the inflection in terms of how much leverage you're going to get on the gross margin side, if we look out, let's say, 3 or 4 quarters?

And I'm trying to understand the comment about how you're going to see a more material inflection in the back half of calendar '20.

So a couple of things.

Let me talk about the dynamics in the current period.

We talked about -- so when we look at where we started from, all end markets in flash were not created equal relative to pricing and margin.

We also noted that mobile was an inferior performing segment for us.

We are taking more of that on this quarter as a percentage of the total.

So that is having a bit of a headwind relative to the flash margin in the current quarter.

So Steve's comments earlier on 2020, we see supply/demand in pretty good shape as we come into 2020, but the normal seasonality of the year, we got to make sure we're managing through that in a cautious way.

And we expect that as the year moves on, as we head towards the middle of the year or the back end, that will continue to improve, and the rate of margin improvement in flash will accelerate in the back half of the year.

Okay.

And then in terms of the NAND gross margin, I mean, does that go back to like 30s in the back half of '20?

Or -- I mean just any sort of rough metric would be helpful.

Well, we are -- I mean let me be clear on this.

We're not providing guidance beyond what we've done in terms of flash gross margins.

But I'll tell you where we need to get to and where we want to get to is back to where flash gross margins are in that 40% range.

That's a margin level that we view as attainable over a period of time, and it's also a margin level that we believe is required to get sufficient return on the capital that we are investing in this business.

So that's where we want to head to.

Our next question comes from C.J. Muse with Evercore.

I guess another question on gross margins.

Specific to the NAND side, can you quantify the K1 fab costs in the September quarter?

I think you said $75 million in the December quarter, and how we should think about that progressing into 2020.

And then I guess as a second question there, on the flash bid side, it looks like -- implied in there, given these costs, roughly bit growth of only 10% or so in end of quarter.

So it looks like you're only growing about 19%, 20% for the year versus many of your competitors who are suggesting low-30s for the industry.

And so I guess is that a function of just deciding not to want to play in the mobility side?

Is it a function of not having the right bits?

Or is there something else?

No.

Let me address the last question first, in terms of us growing.

And first off, the big growth that we're seeing from a demand or revenue perspective is consistent with previous expectations, and it hasn't changed.

And by the way, one of the things that you have to keep in mind is that we took a meaningful amount of production off-line starting earlier this year, independent of the power outage because we saw the oversupply conditions.

Well, we saw it coming and we saw that situation being -- let's just call it untenable.

And so we were reducing our bit output from a supply perspective to help offset that growth -- that unnecessary growth from a supply perspective that created the substantial price decline that we in the industry realized in back half of '18 and also into '19.

So I can ask Bob to address the...

Yes.

So C.J., in terms of that, the September quarter that we just finished, the K1 costs were at $64 million.

And what I said in the guidance as we expect in the December quarter, that will be in the neighborhood of $75 million.

We think they will start to come down from that point in time.

But this is new production capacity we're putting in place.

It becomes a part of our fixed cost over time.

C.J., one other comment I'll make is when you think about this on a bit share comparative basis, when you think about it from a bit capacity output basis, obviously, we continue to remain in the same proximity or ratio to others.

So this was simply us choosing to not produce very low margin product for the reasons Steve said.

Our next question comes from Steven Fox of Cross Research.

Sorry for another gross margin question, but maybe...

We'll have to start banning gross margin questions.

Yes, I know.

Last one.

Let me get this in.

So in terms of just the mix impact on NAND gross margins, can you maybe talk a little bit about what's going on mixed -- how mix is affecting the gross margin guidance for the current quarter versus what you just reported in NAND gross margins.

And then along similar lines, you mentioned some incremental growth next year from things like next-generation gaming and industrial.

Can you talk about their mix impact on gross margins?

And if I dare get in one more, given what you said about the first half of the calendar year, are you able to at least hold gross margins around current levels?

Or based on what you do with -- where you choose to put your bits?

Or do we backtrack a little?

Well, I'm -- again, I'm going to address the last question first.

We would expect that our profitability levels, certainly from a margin percentage standpoint, will continue to modestly improve as we move into the first half of the year.

And back to the mix, let me just reiterate.

Within the flash business, we're taking on a greater portion of mobile business this quarter, which is a margin drag and is dampening our sequential margin performance to some degree on the flash side.

And then Bob also talked about the fab startups.

So those 2 things are affecting the flash margin in this quarter.

When we go into 2020, certainly a number of the new markets are good bit consumers.

The gaming business, we think that, that is going to be not only a good consumer of bits but at reasonably attractive margins.

And industrial and automotive are even better yet, relative to margins and more stable and noncyclical.

Yes.

And Bob already indicated that the K1 costs are kind of a 3-point drag.

And then if you neutralize for mix, I think that we would all be seeing a much more satisfying increase in our flash margins when you neutralize for those 2 items.

And I think that's important for all of us to keep that in mind.

The trajectory is in the right direction.

It's just that there are other moving parts that tend to mask those improving trends.

Got it.

And Steve, congrats on all your accomplishments at Western Digital.

Thank you very much.

Our next question comes from Sidney Ho with Deutsche Bank.

Maybe one more question on this -- on the flash side.

I know you talked about -- a few times about strategically walking away from the mobile mix.

Can you remind us your mix within the NAND flash business today?

And how do you think that would change a year from now?

Is there more tailwind coming from this mix -- change in mix?

And what are the areas that have better margins or worse margins than the average for the business?

So we don't specifically disclose the ratio of flash participation, but I can give you some color on relative performance.

Certainly, the areas of big investment for us that we emphasize are where we want to grow our participation.

So enterprise SSD, fast growing, higher overall margins than average over time.

Industrial embedded both [similarly situation], the higher margins, good growth rates.

And then ultimately, certain segments of mobile over time are attractive and we're focused on those, hence our UFS product investment.

So the higher end, higher performance part of that marketplace.

So we put a big emphasis on quality of revenue and seeking out those higher-margin, more stable end markets and those will be examples that we're focused on.

My follow-up question, I'm sure you'll welcome my hard drive question.

The last earnings call, you talked about introducing 16-terabyte CMR and this 18-terabyte SMR, but you ended up launching a high capacity about 2 months later.

Can you talk about why you made that change in the road map?

And what kind of feedback are you getting from your customers so far?

Yes.

So the reason that we made that decision, it's actually quite simple, is that we made faster progress in terms of our 9-platter platform than what we had previously anticipated.

And so the advancements that we were making from a mechanical perspective allowed us to pull in that platform sooner than what we anticipated, which was favorably received by our customers from a marketplace perspective, which I'll ask Mike to elaborate on.

Yes.

And I think a number of dimensions of this, obviously, it gets them to 18 and 20 terabytes sooner than they would have otherwise.

So the TCO benefit when you combine cost per bit with the other elements like slot tax, because you now -- you eliminate some slots required and also the quality that we deliver with our products.

The whole economic equation is better and then we talked about we get a multiplatform qualification.

So for them, their ability to qualify a single platform that covers all of those configurations is both a simplifying of their efforts, a cost reduction of their efforts and it gets us to market earlier.

So all positive trends from the customers' perspective and getting to that new 18-terabyte CMR and 20-terabyte SMR capacity earlier is a big value.

Our next question comes from Ananda Baruah from Loop Capital.

Steve, congrats as well from me.

I know we have a little more time with you, but it's certainly been fun.

I do have a gross margin question, guys, but I'm going to start with a hyperscale question, just to provide some relief here.

So first, a clarification.

The comment earlier in the Q&A, about 35% growth for calendar '20 year-over-year, was that your hyperscale outlook for calendar '20?

Or was that another...

No, that's -- yes, that's total capacity enterprise.

So hyperscale and OEM consumption of that class of product together.

And can -- could you update for us -- you have an updated view on what this cycle might look like, this hyperscale cycle might look like now.

I think you made some comments over the last couple of quarters.

Yes.

I think the only thing we can say is we continue to see strength of the cycle end of the first half of '20.

I don't think we'd want to go any -- beyond that, that's supported in the 35% year-over-year growth parameters.

So we don't see it rolling over yet.

No.

And the -- any sense of -- if the incremental strength is pulled in this year?

It's been a little bit stronger than we thought.

It's pulled into this year or do your tea leaves point to sustainability as we move -- kind of move into the March quarter?

So we -- it would point to sustainability end of the first half of the calendar year '20.

Okay.

That's great.

And then just quickly on the flash gross margin.

So as we think about the different layers of the headwind that could roll off, it sounds like there's mobile and that 300 -- sorry, the K1 is at 300 basis points, but it sounds like as you participate in more mobile, the headwinds will diminish sort of in coming quarters before we get the inflection in a few quarters.

Will the K1 cost -- so the 300 basis point headwind, will that also kind of gradually roll off until we get into September quarter?

And then what are the other levers that will also contribute to the gradual expansion in margin before we get into the second half of next year?

I just don't want to miss anything.

I want to make sure I'm understanding it appropriately.

Yes.

So I'll start with the K1 question.

So right now, I mean, we're just beginning production in that fab.

And so it will be a gradual improvement in absorption over the next few quarters and eventually, it will be fully absorbed and have a cost structure similar to what we experience elsewhere.

The other drivers are really looking at individual segments.

And as I talked about, we are continuing to expand the product portfolio and increase participation in, let's call it, higher quality of revenue segments like enterprise SSD and other sort of high end mobile marketplaces as we progress into 2020.

So think of it as customer and product mix improvements as well.

Our next question comes from Vijay Rakesh with Mizuho.

Congrats, Steve, and good luck with your next endeavors.

Just on the flash side, I was wondering -- going back on the gross margin, maybe that's the last question here.

If you could look at -- you said you're trying to exit some of the mobile segments, but wondering, as you look at first half, do you expect to step up as you exit some of those mobile segments?

And what kind of a margin improvement should you expect on the flash side with -- as you move more into the SSD or in a less price-sensitive markets?

And the questions on the...

Well, I was going to respond to that, I'm sorry.

But -- so let me clarify, we have not exited from the mobile market.

We chose not to participate in segments of that market for a period of time because the margins -- it was not -- it made no economic sense, okay?

As we see the pricing environment in that market begin to stabilize and improve consistent with improving supply and demand dynamics, we will be increasing our participation in that market starting this quarter.

Got it.

Got it.

And on the hard disk drive side, you got -- you talked about 16 and 20 terabyte.

Can you give us some color of how -- when you expect to ramp some of those products as you go through 2020?

Yes.

So obviously, we said we're sampling those in the current quarter and we expect to ramp that in the first half of calendar '20, so that's in the not-too-distant future.

And our final question today will come from Nehal Chokshi with Maxim.

Just for edification purposes here or yes, education purposes, when does the NAND flash industry typically seasonally peak in terms of bit demand from a monthly perspective?

Well, I mean -- well, this will sound like a facetious answer.

It's the old adage that the best quarter of the year is October, November and December.

It's around that time, but I don't know literally which month, yes.

Okay.

So Mike, I think you made a comment that the exabyte loss from the power outage turns out to be 4 exabytes instead of 6 exabytes, but you still expect that the excess inventory industry-wide would be flushed out by end of this year.

So given that lower exabyte loss, have you seen better demand than expected over the past 3 months?

Yes.

So I think in general our bit consumption and demand has been at or above what we thought.

So our comments where we thought by the end of the year the industry inventories would be substantially improved and we stand by that.

And what's been the driver of that better-than-expected demand for the NAND flash bits?

I think we've seen, generally speaking, some -- a little bit better mix in terms of capacity per unit.

And I think some of the end-markets are a little more robust than maybe we originally expected, including mobile.

Speakers, I'm showing no further questions in the queue.

I'd like to turn the call back over to management for any closing remarks.

All right.

Thank you for joining us today.

I would also like to extend a thank you to all of our employees, customers and business partners.

We look forward to a successful year together.

Have a great rest of the day.

This concludes today's conference call.

Thank you for joining.

You may now disconnect.