Robert Blair - Western Digital Corp. Stephen D. Milligan - Western Digital Corp. Michael D. Cordano - Western Digital Corp. Mark P. Long - Western Digital Corp..

Aaron Rakers - Stifel, Nicolaus & Co., Inc. David Ryzhik - Susquehanna Financial Group LLLP Jeriel Ong - RBC Capital Markets LLC James Kisner - Jefferies LLC Steven Fox - Cross Research LLC Rich J. Kugele - Needham & Co. LLC Mark Moskowitz - Barclays Capital, Inc. Karl Ackerman - Cowen & Co. LLC Ananda P. Baruah - Brean Capital LLC.

Good afternoon and thank you for standing by. Welcome to Western Digital's First Quarter Fiscal 2017 Conference Call. All participants are in a listen-only mode. Later we will conduct a question-and-answer session. As a reminder, this call is being recorded. I now will turn the call over to Mr. Bob Blair. You may begin..

Thank you. And good afternoon, everyone.

This conference call will contain forward-looking statements within the meaning of the federal securities laws, including statements concerning our expected financial performance for our second fiscal quarter ending December 30, 2016, our market positioning and growth opportunities, our strategic platform, integration activities and achievement of synergy goals, demand in market trends, the benefits of our short-term incentive compensation programs, our product portfolio, product features and their benefits, product transitions and other development efforts and customer acceptance of our products and manufacturing yields and cost improvements.

These forward-looking statements are based on management's current expectations and are subject to risks and uncertainties that could cause actual results to differ materially, including those listed in our annual report on Form 10-K filed with the SEC on August 29, 2016.

We undertake no obligation to update our forward-looking statements to reflect new information or events. Further, references will be made during this call to non-GAAP financial measures.

Reconciliations of the differences between the historical non-GAAP measures we provide during this call to the comparable historical GAAP financial measures will be posted in the Investor Relations section of our website.

We have not reconciled our non-GAAP financial measure guidance to the most directly comparable GAAP measure because material items that impact these measures are out of our control or cannot be reasonably predicted.

Accordingly, a reconciliation of the non-GAAP financial measure guidance to the corresponding GAAP measures is not available without unreasonable effort. In the question-answer session part of our call today, we ask that you limit yourselves to one question and one follow-up to allow as many callers as possible during our allotted time.

I thank you in advance for that. And I will now turn the call over to our CEO, Steve Milligan..

Good afternoon and thank you for joining us. With me today are Mike Cordano, President and Chief Operating Officer, and Mark Long, Chief Financial Officer. After my opening remarks, Mike will provide a summary of recent business highlights and Mark will cover the fiscal first quarter results and wrap-up with our guidance for the fiscal second quarter.

I'm pleased with our performance in the September quarter, our first full quarter as an integrated company following the SanDisk acquisition in May. We reported revenue of $4.7 billion, non-GAAP gross margins of 34%, and non-GAAP earnings per share of $1.18, exceeding our revised guidance in each category.

We continued to experience strong demand throughout the September quarter for both hard drive and flash-based products from all categories of customers, largely driven by cloud and mobility based applications, as well as better than expected PC demand and our own outperformance in certain categories.

These trends are expected to continue in the December quarter. We are delivering our financial performance through continued strong execution by the newly-combined Western Digital team, demonstrating the power of our broad-based and unique platform.

The platform includes an unparalleled breadth of storage product offerings, expanded vertical integration and go-to-market capabilities, an increasing level of customer intimacy and a proven ability to innovate and sustain leadership across multiple storage technologies.

We have made significant progress establishing this strategic platform by combining and leveraging the collective strengths and resources of legacy HGST, SanDisk and WD. Our platform is aligned with our vision of the continued rapid growth of data and its increasing importance in our everyday lives. The power and promise of data is undeniable.

It is creating a world that's more predictive, more productive and more personal, enabling smarter decisions, breakthrough discoveries and deeper connections. The Western Digital platform is at the core of enabling the possibilities of data.

It differentiates us from our competitors and we believe it will drive our company to a future of growth and long-term industry leadership. We are already seeing the benefits of the platform in our differentiated financial results and outlook.

I am very pleased that we have achieved this financial performance, while, at the same time, successfully executing on two top priorities, namely, the integrations of HGST, SanDisk and WD, and the transition to next-generation NAND technology. Integration of the three companies is proceeding very well, and our synergy achievements are on track.

I'm also pleased to note that our transition to 3D NAND technology continues to progress as planned and we will begin a significant commercial ramp of our second-generation 3D NAND with 64 layers in the first half of calendar 2017, as previously announced. Mike Cordano will now report on our business highlights in the September quarter..

Thank you, Steve, and good afternoon, everyone. I'm pleased with our operational execution and the resulting business performance for the September quarter. The industry supply/demand environment in both HDDs and flash was favorable.

With a full quarter of the legacy SanDisk business results integrated into the company, our Q1 report reflects the strength of our broad portfolio addressing a similarly expanded market opportunity.

As part of our mission to build a compelling Western Digital platform, we continue to make progress with integrating the HGST and SanDisk businesses and we are on pace with our plan to further streamline our product and technology base (7:06). Turning to our highlights in the quarter.

In data center devices and solutions, we delivered solid year-over-year growth in total exabytes shipped, primarily driven by continued strength in cloud-related demand for our near-line capacity HDDs, with that product category alone driving almost 50% year-over-year growth in exabytes shipped.

I'm very pleased to report that our industry-leading 10 terabyte helium drive gained further adoption at several of our major OEM and cloud customers, reflecting the compelling value proposition of this offering.

Just two weeks ago, we achieved an important milestone for our HelioSeal platform, with cumulative shipments exceeding 10 million units, representing approximately 76 exabytes of storage capacity since the launch of this platform in late 2013.

This technology, now in its third generation, enables even higher density HDDs specifically designed for data center applications and will be able to incorporate future magnetic recording technologies.

In data center solutions, we are pleased with SoftBank's adoption of our InfiniFlash all-flash storage platform for their new software-defined storage solution. In client devices, we experienced strong demand for our HDDs and SSDs for PCs, as that market continued to show signs of stabilization and we outperformed in both product categories.

Our 3 bits per cell based embedded NAND solutions saw increased adoption in mobile phones, as this technology offers a compelling value proposition compared to 2 bits per cell based alternatives. I am pleased to note that our embedded 3 bits per cell solutions are now qualified and shipping to the majority of the world's top mobile handset vendors.

We are now deepening our participation in the automotive, connected home and industrial categories, reflecting our commitment to driving innovations beyond the traditional mobile application. These are important long-term growth opportunities with longer product cycles for both our embedded and removable offerings.

Our performance in client solutions, which includes retail offerings from legacy HGST, SanDisk and WD, was strong during the September quarter, and our combined portfolio of flash and HDD products has generated excellent reception from customers.

Following a successful back-to-school sales period, performance in our retail channel remains solid, reflecting strength in both demand trends as well as preference for our brands. We demonstrated continued innovation in our removable product categories with the announcement of the world's first 1 terabit SD card.

From a NAND technology perspective, we have begun OEM sampling of our second generation of 3D NAND technology of 64 layers and we're on track to deliver volume shipments of removable product with this technology in the December quarter.

Our previously discussed plan to begin a significant commercial ramp of 64-layer 3D NAND in the first half of calendar year 2017 remains on track. Our first generation of 3D NAND technology with 3 bits per cell in a 48-layer architecture is shipping in embedded mobile and removable products.

We continue to ship products using our 15 nanometer 2D NAND technology, which is achieving new manufacturing milestones in terms of yields and cost improvements. In summary, I'm pleased with our first quarter performance and the execution by our team.

The opportunities we are creating through integrating HGST, SanDisk and WD are significant and we are making good progress in setting up Western Digital for long-term growth and success. In the December quarter, we expect overall supply/demand conditions in both HDD and NAND to remain constrained resulting in favorable industry dynamics.

The longer-term drivers of demand for storage are vibrant and we are focused on our offering industry-leading portfolio to drive the best results for our customers and shareholders. With that, I will turn the call over to Mark for the financial discussion..

Thank you, Mike. I'm encouraged by our financial performance this quarter. Our teams executed well in a solid market environment as we capitalized on our strong product offering, achieved targeted costs and efficiency improvements and reduced our cost of debt, primarily due to two re-pricing transactions.

Our key financial metrics continued to improve following the revised guidance we provided on September 7 due to both favorable market dynamics and consistent execution by our team. Our revenue for the September quarter was $4.7 billion, driven by strong performance across each of our end-markets.

Revenue in datacenter devices and solutions was $1.4 billion, client devices was $2.3 billion, and client solutions was $1 billion. Our datacenter revenue growth continues to be driven largely by cloud-related storage demand.

As a result, we saw continued strength in capacity enterprise HDDs, stable demand for performance enterprise HDDs, as well as positive market dynamics for enterprise SSDs. Client devices benefited from seasonally strong demand for HDDs in gaming and PC applications as well as flash-based products in PCs and mobile handsets.

In client solutions, our revenue grew as a function of seasonality and healthy demand for removable and other flash-based products. Our non-GAAP gross margin was 33.6%, which was up 260 basis points versus the June quarter.

We were able to achieve this margin expansion through continued product cost improvements and better than anticipated pricing for certain HDD and flash products. Our product cost improvement resulted from increased volumes, consistent execution and ongoing progress with our integration activities.

Gross margin also benefited from our first full quarter of flash revenue, which helped to offset the seasonal increase in lower margin HDD gaming revenue, the latter of which typically peaks in the September quarter. Turning to the discussion of operating expenses. Our non-GAAP operating expenses totaled $952 million.

We incurred higher short-term incentive compensation due to stronger business performance, which nonetheless was partially offset by continued progress towards our integration synergy target. Let me take a few moments to explain our short-term incentive compensation program. It is based on goals set on a semiannual basis.

We believe this semiannual approach to incentive compensation allows us to effectively navigate changing market environments. To understand the underlying business performance, it is useful to normalize these payments across periods.

For example, the payouts in the first six months of calendar 2016 were significantly below target, while the payouts for the second half are currently expected to be above target. As a result, the annualized payout would be at a roughly 100% target level. Non-GAAP interest and other expense for the quarter was $227 million.

On August 17, we reduced our U.S. Term Loan B balance by $750 million through a voluntary prepayment and repriced the remaining $3 billion balance. On September 22, we repriced our euro-denominated Term Loan B with a balance of €885 million.

The aggregate annual non-GAAP and cash interest savings from the two repricing transactions are approximately $150 million and $120 million, respectively, on a going forward basis. These transactions resulted in $267 million of debt extinguishment charges in our GAAP interest and other expenses, $227 million of which were non-cash charges.

Our non-GAAP effective tax rate for the September quarter was approximately 16%, within our expected range of 15% to 20% for this period. On a non-GAAP basis, net income in the September quarter was $341 million, or $1.18 per share. On a GAAP basis, we had a net loss of $366 million, or $1.28 per share.

The GAAP net loss for the period includes charges associated with our recent acquisitions, debt extinguishment charges related to our repricing and repayment of outstanding debt. Essentially, the entire net difference between our GAAP and non-GAAP net income is a result of non-cash charges.

In the September quarter, we generated $440 million of cash from operations, with $210 million spent on capital investments resulting in free cash flow of $230 million. We also declared a dividend in the amount of $0.50 per share.

We closed the quarter with total cash and cash equivalents of $4.1 billion, compared to the June quarter end balance of $8.2 billion. The decrease was primarily driven by repayment of $4.2 billion of debt, consisting of the acquisition bridge loan, U.S. Term Loan B and SanDisk convertible debt.

As Steve and Mike indicated, we have continued to make very good progress with respect to our integration.

From a synergy standpoint, we remain on track to achieve the $800 million of annualized savings from the HGST integration by the end of calendar 2017, and expect to exit the December quarter of this year having achieved more than $175 million of cost of revenue synergies and approximately $300 million of operating expense synergies each on an annual run rate basis.

With respect to the SanDisk integration, we expect to exit the December quarter having realized synergies of approximately $130 million on an annual run rate basis toward our 18-month target of achieving $500 million of total run rate synergies on an annualized basis.

Now, I would like to describe a change we are implementing in our non-GAAP reporting, beginning with our December quarter.

To better reflect the performance of the underlying business operations in our results, and to be more consistent with the majority of our technology peers, we will be excluding stock-based compensation expense from our non-GAAP results. All future guidance, including today's for the December quarter, will be on this basis.

The stock-based compensation portion of our total operating expenses is typically between $75 million and $85 million per quarter.

If we had made this change in our September quarter, our non-GAAP operating expenses would have declined $89 million from $952 million to $863 million, our non-GAAP cost of revenue would have also decreased $15 million and our non-GAAP tax expense would have been reduced $3 million.

The total stock-based compensation expense impact would have increased non-GAAP net income by $107 million to $448 million. Non-GAAP EPS would have increased from $1.18 to $1.54. We are posting this data with some historical comparisons on the Investor Relations section of our website for your information.

With that, I will now provide our guidance for the December quarter on the new non-GAAP basis, excluding stock-based compensation expense.

While the September quarter is typically the high point of our revenue due to seasonality, we expect our December quarter revenue to be equally strong due to the continuation of favorable industry dynamics and the success of our broadened product portfolio.

As such, we now expect revenue in our December quarter to be approximately flat to the September quarter. On a non-GAAP basis, we expect gross margin to increase to approximately 35%, driven by improved product mix reflecting lower gaming volume and outperformance in other key areas.

I would like to note that we have only included half a quarter of license and royalty revenue from our cross-license agreement with Samsung, which, as you know, expired on August 14 and under which revenue was recognized one quarter in arrear.

We remain in constructive negotiations with Samsung, working toward a new agreement and will provide further updates as this situation changes. Turning to operating expenses.

We expect those to total approximately $805 million, which includes higher variable compensation expense given our outperformance of the targets set for the first six months of our fiscal year.

Interest and other expense is expected to be approximately $205 million, reflecting a full quarter benefit from our lower interest expense and amortization of debt issuance costs from the debt repricing and paydown transactions. We expect an effective tax rate in the 14% to 16% range.

As a result, we expect non-GAAP earnings per share between $1.85 and $1.95, with an estimated share count 293 million diluted shares. I would note, on our previous non-GAAP basis including stock-based compensation expense, the high end of our earnings per share guidance would have been approximately $1.60.

At our Investor Day on December 6, we will provide a deeper dive into our strategy, business operations, technology capabilities and long-term financial model, which should give further clarity on what we expect to deliver across key financial metrics for the new Western Digital. I will now turn the call over to the operator to begin the Q&A session.

Operator?.

Ladies and gentlemen, we will now begin the question-and-answer portion of today's call. Our first question comes from the line of Rod Hall with JPMorgan. Your line is open. Please go ahead..

Hi. This is (23:19) on behalf of Rod. Thanks for taking my question. My first question is on your revenue beat. I see that your net revenue is 4% ahead of your pre-announcement, which only happened last month.

So what kind of linearity trends you are seeing? Did the market trends accelerate towards the back half of the quarter?.

Sure. So for the quarter, linearity was very strong. And we experienced strong demand and favorable pricing since the guidance we provided in early September..

Yeah. So if you look at it, of our revenue over-achievement from our previous guidance of $4.5 billion to the $4.7 billion, about 80% of that over-achievement was due to higher than expected demand and the balance was a bit better pricing..

Okay. Great.

As a follow-up, could you comment on your expectations for the NAND market bit growth as well as your bit growth? And then a bit on your expectations for pricing?.

Yeah. So talking about the current calendar year, we would expect that bit growth to be around the mid-30%, about 35% for the industry. And in our case, we had previously talked about something in the low-30% induced to some very good work by our operational teams that will be closer to the industry rate exiting this year.

We'd expect next year to be in the mid-40%s, approximately 45%, both for the industry and for ourselves as we transition as an industry and we transition to 3D NAND technology..

Okay. Great. Thanks..

Thank you. And our next question comes from the line of Aaron Rakers with Stifel. Your line is open. Please go ahead..

Yeah. Thank you for taking the questions and congratulations on the quarter.

As we think about building our models and you look at the variable compensation components that are included in the expectations, how should we think about variable comp going into the December quarter? And I think at one conference, you had recently noted that that will start to tail off as you realize some of the synergies from these integrations, particularly HGST..

So the variable comp that is really driving the deltas here is our short-term bonus that's paid over a six-month horizon, as I was indicating earlier. And so for Q4, calendar Q4, we'll still have some of that effect. But as you go into calendar Q1, our fiscal Q3, that effect will be eliminated as we reset the target.

And our forecast would include a normalized 100% target achievement there. So the best way to think about this normalized OpEx is that, from a calendar Q1 standpoint through to really the end of fiscal 2017, you're going to see a steady trend downward. So, first, when we talk about normalizing, we exclude the stock-based compensation.

And then we bring our short-term bonus to 100%. So for the first six months of the calendar year, as we said, you'll actually be adding back some of the bonus that we didn't pay because we were below that 100% target.

And for the remainder of the calendar year 2016, you would be deducting that overage, and then for the first half of calendar 2017, you would not have that short-term bonus effect..

Okay. That's helpful. And just as a follow-up, I'm just curious on cloud pretty universally has been cited as strong across many different companies.

I'm curious beyond the December quarter how you're currently thinking about cloud demand, near-line demand, in particular going into the first half of 2017?.

Yeah. Let me just follow up and give you another data point because we're conscious of the challenges modeling some of the transitions here. But when you think about the first quarter of calendar 2017, I want to kind of give you a perspective on where the normalized OpEx should range.

And a way to think about that for modeling purposes would be in the $770 million range. So, hopefully, that helped..

So, we wanted to close off that, Aaron, and Mike can talk about cloud demand..

Perfect. Very helpful..

Sure. All right. Relative to the cloud demand capacity enterprise, specifically I noted in my prepared remarks, we saw a year-over-year increase of nearly 50%. We see this year in total running around 38% or above our long-term model of 35%. As we look into calendar 2017, we're not updating our long-term model at this point.

But, we would suggest that things will run to sort of a bias up above that kind of consistent with what we're seeing this year..

Okay. Thank you. Good luck..

Thank you. And our next question comes from the line of Mehdi Hosseini with Susquehanna. Your line is open. Please go ahead..

Mehdi Hosseini, thank you..

Yeah, actually it's not Mehdi, it's David Ryzhik for Mehdi Hosseini on the line. Thank you so much for taking the question.

Can you guys talk about your HDD capacity in terms of units and/or exabytes? And what you've done in that regard and how that trend over the next quarters? And in that context, how we can think about capacity utilization? And I have a follow-up..

Yeah. So, this is Steve, I'll take that. So I would say that, I mean, there's a lot of dynamics that are going on in terms of our hard drive capacity because as you know, over the last kind of several years, we've been going through a manufacturing rationalization process given declines that we've seen in the hard drive market.

So, there are kind of different puts and takes when you look at various different products for various different factory locations. But on balance right now in terms of hard drive manufacturing, I would say that we are kind of reasonably balanced. In other words, we're able to largely meet overall customer demand.

And again, we're going to continue to rationalize this as we move forward. The one thing that we're a little bit cautious about and keeping our eye on is that, obviously when you're talking about capacity enterprise products, which, as Mike indicated, the demand in that space has been reasonably robust.

And that uses a lot of heads and disks because of the platter count. And we have less flexibility both in terms of lead time and capacity in terms of heads and disks.

And so as we look at that and as we watch and work closely with our customers, if for one reason or another we would see, call it, a spike from a demand perspective, our ability to meet some of that upside might be more limited. So, that's one thing that we're a bit cautious on right now.

But overall, we feel pretty good about where we're at from a utilization and capacity perspective on the HDD front..

Great. Thanks, Steve. And as a follow-up for Mark, how can we think about the cash conversion cycle, and more broadly speaking, like a normalized free cash flow rate as you've gotten your arms around the balance sheet and cash generation capability? Just any insight in how we can model that going forward. Thanks..

Sure. So from a cash conversion cycle standpoint, I think you can model basically where we've been over the last couple of quarters, but you'll have some changing dynamics. So what I would expect is inventory will likely trend up and this is really a function of us building strategic reserves as we restructure parts of our supply chain.

And then eventually, you'll see the NAND parts of our business, the allocation will begin to soften and we'll begin to build more inventory. So, that will trend upwards a little bit, but you'll then I think see payables be extended from where they were in Q1. So, I think, overall, you'll see a cash conversion cycle in about the range that it's been.

And in terms of the cash flow generation margin, we're actually going to have a little bit more noise in that equation for the next few quarters, primarily as a function of our restructuring and ongoing integration activities.

But, one of the things we will be doing during Investor Day is providing more visibility into the dynamics of that as well as a way to think about our long-term financial modeling and the key metric around cash flow generation..

Great. Thanks, Mark..

Thank you. And our next question comes from the line of Amit Daryanani with RBC Capital Markets. Your line is open. Please go ahead..

Hey, this is Jeriel on behalf of Amit. I just have a few questions here. Number one, can you just touch on what drove the gross margin upside? I think it was up about 60 basis points versus your positive pre and 160 basis points versus initial guide.

Was this more HDD or NAND or cost integration, and some rank order would be valuable?.

So I think as we mentioned, the favorable pricing was across both parts of our business, but primarily in the NAND area. And we did experience some additional benefits from mix, especially given the fact that we had the first full quarter of the flash product revenue..

Yeah. And also the strong performance in terms of capacity enterprise as well..

And I mean, there's a lot of dynamics. The other thing I'd just add is that we continue to make progress in terms of our integration synergy goals. So, that continues to provide some headwind – or not headwind, tailwinds for us in terms of our margin performance.

So, it was really kind of oversimplified, it was kind of a little bit of everything, kind of across the board..

All right. Thanks for that..

Yeah. Just to finish that thought, we would expect those same drivers to carry forward into the second fiscal quarter and that's driving the expansion of gross margin..

Thank you. And as a follow-up, I understand that you're only including about 45 days of that Samsung contribution, which makes sense. If or when you get a resolution with Samsung, would they back pay your payments from November 15 onwards or....

We would certainly hope so, yes, because the agreement expired on August 14. So when we put a new agreement in place, we would expect it to be retroactive to that date..

All right. Thank you so much..

Thank you. And our next question comes from the line of James Kisner with Jefferies. Your line is open. Please go ahead..

Thank you very much. So I believe you said near-line demand was up 50% year-over-year in exabytes shipped. I was wondering if you could just comment a little bit on how that translates to revenue change year-over-year, and I guess, related to that, change in price per bit year-over-year.

And I guess relatedly, have you seen a more rational competitive environment for near-line and what has happened to your share in near-line quarter-on-quarter? Thanks..

So from a translation to a revenue standpoint, it's about half that in terms of the increase in revenue..

Yeah. In terms of market dynamics, I think we think it's fairly balanced and as we expected. And I think relative to total bit participation, we think we've improved our position sequentially and we see a trend of that going forward really relative to our increasing sales of our 10 terabyte offering..

All right. Thank you very much..

Thank you. And our next question comes from the line of Steven Fox with Cross Research. Your line is open. Please go ahead..

Thanks. Good afternoon. Just getting back to your revenue guidance for a second, I was wondering if there's any other puts and takes you can talk to in terms of how you get to the flat, either by the three business segments you've established or maybe even hopefully by some product categories as well? And then I had an enterprise follow-up question..

Sure. So let me talk about our end-markets. For datacenter devices, you can think about that whole category being roughly flat. When it comes to client devices, I think within that, while the whole category again will be roughly flat, you'll have some puts and takes.

You'll see continued strength in client SSD, but we'll see gaming and, to some extent, notebook HDDs sort of come off that seasonal peak and then we'll see a little strength in desktop. When it comes to our client solutions, we actually see strength across both our HDD and flash-based products.

So I think on balance, it's flat to slightly up for each of the three categories..

The largest single variable there is that seasonality of the gaming marketplace, so that peaks, as Mark said, in calendar Q3. So that's the largest single variable within that..

Great. That's helpful. And then just as a follow-up, as you mentioned, you have a unique view into enterprise storage now. I was wondering if you could just compare average capacity trends between HDDs and SSDs into some of the cloud data centers that you have projects in and how the growth rates compare and contrast. Thank you very much..

Yeah, on a growth rate basis, the capacity growth rate for flash exceeds that of disk. And so that trend continues. That's been ongoing for some time and that's really specifically related to hyperscale deployments..

Thank you very much..

Thank you. And our next question comes from the line of Rich Kugele with Needham & Company. Your line is open. Please go ahead..

Thank you. Good afternoon. Just one question for me, tagging onto that last question. So Micron, for example, has been touting its 15 terabyte drive.

And I'm just interested in, may be best from Mike, how do you see the next few years playing out on the highest capacity points between hard drives and SSDs? How do you position these two? And does WD get to a point where it just is indifferent to whatever people select for cloud hyperscale environments? Thanks..

Yeah. So I think, Rich, it still begins with the workload and then the cost per bit that optimally supports that workload. So going forward, we don't see a convergence to an all-flash world anytime soon, frankly, anytime in our planning horizon. So what we do see is continued growth of capacity enterprise that we articulated.

That's going be long into the future, although flash revenue and flash bits on a proportionate basis is growing faster. So from that standpoint, that's the trend we would point you to. And they have very different purposes in the hierarchy.

So we continue to obviously work closely with those and are innovating around data center architectures and we intend to deploy our technology and our products in an optimal way to those trends..

How much of a price differential is there today between the SSD version and a comparable capacity hard drive?.

It varies depending on performance requirements, et cetera, but it would run between 5X and 10X..

Okay. Great. Thank you..

Thank you. Our next question comes from the line of Mark Moskowitz with Barclays. Your line is open. Please go ahead..

Yes. Thanks. Good afternoon. I apologize if the questions have been asked. I've been jumping between a couple calls.

The first question is, can you talk a little more about next year if demand improvement in both flash and disk drives continue from an end market perspective, could we see any capacity cuts driving a shortage or a constraint issue next year from one or both segments? And the other question is around the exclusion of stock options.

Can you talk a little more about the philosophy there, just given that we were aware of SanDisk and the overhang there previously, why the change now and what's the impetus? That would be great. Thank you..

Okay. Let me talk about the supply/demand environment. We said publicly and we'll reiterate here that we see constrained supply in the NAND side of our business through calendar year 2017. And that's primarily driven by two things, end market demand for that class of technology and the rate and pace of the industry's conversion to 3D NAND.

Relative to the hard drive market, I think at this point we would see that as a normal or balanced environment..

And then, Mark, with respect to the decision to exclude stock-based compensation expense from our non-GAAP results, we really had two primary drivers. One was just the way we run the business. And our stock-based compensation expense actually has some elements of it that are either short term in nature or are variable.

And as a result, it injects some variability that undermines our ability to demonstrate or track the improvements we're making from a synergy realization standpoint. So that's the first driver.

The second driver is that we are more and more focused on our cash flow generation and looking at our results from a cash standpoint, particularly given our balance sheet. And by excluding the stock-based compensation, we are better able to highlight and track the cash flow impact of our operating decision..

Now, one thing, Mark, that I will add to that. This is Steve. Is that we fully recognize the fact that there is an impact in the form of dilution from stock-based compensation and we remain not only aware of that, but sensitive of that.

And if you look at our proxy statement in terms of compensation philosophies and that sort of thing, there's no change in that regard in terms of what we're doing. It's more a matter of excluding those figures to provide more meaningful analytical information, if you want to call it that, to the financial community.

But I don't want people to think that we don't recognize the fact that there is a dilution component as it relates to our existing shareholders..

And Mark, I think as you know, part of the calculus was looking at the comparable peers. And what they do to try to provide an easier basis for comparison, so..

Okay. Thank you..

Thank you. And our next question comes from the line of Karl Ackerman with Cowen & Company. Your line is open. Please go ahead..

Hi. Thank you for taking my questions. I know it's difficult to guide NAND gross margins since ASPs are beyond your control.

But do you think your cost improvement will be above or below your long-term expectations of 20% in 2016 as you ramp 3D? And maybe, what's your expectation in 2017, calendar 2017, as 3D becomes a larger part of your mix? And I have a follow up, please..

Okay. So we've talked about sort of the cost reduction domain of kind of three generations – or three periods in the NAND industry. Very early on, it was running at around a 50% annualized rate. The next period was running at about a 30% rate. Frankly, that's now ending as we come to the conclusion of plainer NAND.

As we move fully as an industry into 3D, once we're through the transition, we expect around a 20% per year cost reduction capability. So, that's the way we would guide you to think about that. Obviously, there are some perturbations as we transition from 2D to 3D, but that's the way to think about it from a long-term modeling perspective..

Got it. I can appreciate that.

And then in the context of your bit growth outlook for calendar 2017 that you just gave and also your past comments that 3D NAND will be 40% of your overall bit output by the end of next calendar year, should we expect the majority of your 3D outlook to be 64-layer or will it be more weighted toward 48-layer?.

It will be 64-layer.

And I think we have – what's the right number for exit rate in 2017?.

It's around 40%, but its wafers out..

Oh. Okay. So not bit output, it's just wafers out? Okay. So the 40% is wafers out..

Correct. And the other comment I'll make, just to clarify, 48-layer, we've talked about, is really a learning node for us. So, to just reiterate what Steve said, the volume as we exit next calendar year will be predominantly the 64-layer..

Thank you very much..

Our next question comes from the line of Ananda Baruah with Brean Capital. Your line is open. Please go ahead..

Hi, guys. Hey, congratulations and thanks for taking the question. One and then a quick follow up.

Steve, would love your, I guess, your context on given that on the hard drive side of things, the capacity in the industry is pretty well rationalized, and I guess, has the potential based on cloud growth going into next year to maybe be, it sounds fantastic, but close to 100% by midyear.

What implications, if any, does that then have on, I guess, the context, the tenor of your relationships with customers, your conversations with customers, business practices with customers in any regard? Would love to get your sense there and then I have a quick follow up on a different topic..

Well, yeah, and Mike can help me with this as well, but I think that one of the things that has really happened over a period of time, because as the hyperscale guys grow, and they get bigger, and as they need to add capacity and their growth is accelerating. And so you just get into, call it, the law of large numbers.

We have to do a much better job working with the hyperscale guys on longer-term capacity planning. And so, we believe we've done a good job on that because we don't – they don't want to ask us for product and us not have it. And then, clearly, we don't want to gap out on them.

And so, it's driven in effect a much more close-knit planning process, longer-term planning process, which the good side of that is that it allows for a little bit more predictability in terms of what's going on in that side of the business, because there was a time – well, it can still happen.

But there's been conversations about the lumpiness of the business and that can still occur. But because of the dynamics that we're seeing, we're actually getting into a little bit more longer-term planning and a little bit more predictability in terms of what's happening from a demand perspective. Mike can kind of add to that..

Yeah, and I'll just sort of reiterate some of the reasons this is. So, we've talked about this long-term growth rate of 35%. We've been, as an industry, achieving that with a number of efficiency factors flowing through. We've highlighted in the past both technological as well as operational.

So, technologically, things like erasure coding are widely deployed now in that infrastructure. So, the sort of efficiency gains technologically, although we expect some in the future, the largest magnitude of that has sort of run its course. The other one is operational efficiency and that underpins what Steve said.

They're maturing as operational units, their ability to sort of do things like manage inventory and repurpose and reutilize their fleet. Those processes are maturing.

So, really, what we now see and perhaps one of the reasons we're seeing a bias up on the exabyte growth rate is we're kind of working through those things and we're starting to approach a more natural storage deploy requirement. So that's something we'll continue to monitor and see how that trends.

At this point, we'll kind of stick with 35%, but as I said, I think our bias would be up from there going forward..

And Steve and Mike, was sort of working more closely with those folks, would that include anything in the context of contractual alterations? Could a long-term – I'm just coming up off the top of my head, long-term contracts play into it in any context, maybe capital sharing play into it in any context, anything with financial implications?.

Well, I'll say this. I think the general view of us now as we put together this broader set of assets is we're more relevant, we're more strategically important, and our engagement is changing with the broader marketplace. So, things along those lines and other strategic things are all sort of more possible now..

Got it. Thanks so much..

Thank you. So, I want to thank you all for joining us today. We look forward to seeing many of you at our Investor Day on December 6. Have a good day..

Ladies and gentlemen, this does conclude the program, and you may all disconnect. Everyone, have a great day..