Jul 17, 2013
Executives
Brendan Lahiff Sanjay Mehrotra - Co-Founder, Chief Executive Officer, President, Director and Member of Special Option Committee Judy Bruner - Chief Financial Officer, Executive Vice President of Administration and Member of Secondary Executive Committee
Analysts
Michael Chen Mehdi Hosseini - Susquehanna Financial Group, LLLP, Research Division Brian C. Peterson - Raymond James & Associates, Inc., Research Division Earl Hege - RBC Capital Markets, LLC, Research Division Mark C.
Newman - Sanford C. Bernstein & Co., LLC., Research Division James Schneider - Goldman Sachs Group Inc., Research Division Vijay R.
Rakesh - Sterne Agee & Leach Inc., Research Division Daniel L. Amir - Lazard Capital Markets LLC, Research Division Michael Bressler Gabriel Ho Craig A.
Ellis - B. Riley Caris, Research Division Steven Bryant Fox - Cross Research LLC Harlan Sur - JP Morgan Chase & Co, Research Division Stephen Chin - UBS Investment Bank, Research Division Dean Grumlose - Stifel, Nicolaus & Co., Inc., Research Division
Operator
Good day, everyone, and welcome to the SanDisk Corporation Second Quarter 2013 Financial Results Conference. Today's conference is being recorded.
At this time, I would like to turn things over to Mr. Brendan Lahiff with Investor Relations.
Please go ahead, sir.
Brendan Lahiff
Thank you, and good afternoon. With me on the call today are Sanjay Mehrotra, President and CEO of SanDisk; Judy Bruner, Executive Vice President of Administration and CFO.
Before we begin, please note that any non-GAAP financial matters discussed during this call as defined by the SEC in Regulation G will be reconciled to the most directly comparable GAAP financial measure. That reconciliation is now available, along with supplemental schedules on our website at sandisk.com/ir.
Please note that non-GAAP to GAAP reconciliation tables for all applicable guidance will be posted on our website. This guidance is exclusive of any one-time transactions and does not reflect the effect of any acquisitions, divestitures or similar transactions that may be completed after July 17, 2013.
In addition, during our call today, we will make forward-looking statements. Any statement that refers to expectations, projections or other characterizations of future events, including financial projections and future market conditions is a forward-looking statement.
Actual results may differ materially from those expressed in these forward-looking statements. For more information, please refer to the risk factors discussed in the documents we file from time to time with the SEC, including our annual report on Form 10-K for fiscal 2012 and our subsequent quarterly reports on Form 10-Q.
SanDisk assumes no right -- no obligation to update these forward-looking statements which speak as of today. With that, I will turn the call over to Sanjay.
Sanjay Mehrotra
Thank you, Brendan, and good afternoon, everyone. We are pleased to report record second quarter revenue and strong margin.
We made excellent focus with our embedded mobile product portfolio with new design wins, and our SSD business gained further momentum. We remained focused on strengthening our product portfolio, and our execution in technology transitions remained outstanding.
Our second quarter achievements have put the company on a path to deliver the record 2013. Our second quarter performance was driven by strong growth in our mobile embedded products with iNAND and iNAND MCP, causing the largest revenue gains during the quarter.
Our embedded solutions team has made significant progress over the past year, and their efforts to diversify our customer base with an expanded array of compelling solutions is bearing fruit. Both smartphone and tablet adoption rates continue to be strong in the industry with industry analysts forecasting more than 900 million smartphones to ship in 2013, with tablet volumes expected to approach 200 million.
Further, China smartphone OEMs are becoming increasingly important in the global smartphone market, with analysts estimating the top China brands to ship more than 10% of smartphone volume this year. SanDisk has made significant inroads into this growing market with our iNAND and iNAND MCP offerings, building further upon our strong presence with global smartphone OEMs.
New smartphone launches, including numerous premium brands, continue to feature card slots to expand device storage beyond the embedded capacity. In particular, the attach rates of slots to non-iOS phones remains greater than 90%, indicating strong OEM interest in providing storage expansion options to the end customer.
This trend continues to play in our favor as reflected by another strong quarter for our retail business, which was in part driven by year-over-year and sequential revenue growth from retail sales of our branded microSD cards. Recently, we introduced the world's fastest 64-gigabyte SanDisk Extreme microSD card.
With read and write speeds of 80 and 50 megabytes per second, respectively, the SanDisk Extreme microSD card further extends our high-performance mobile product offering to address increasingly features such as camera and video applications in not only smartphones and tablets but in other device categories, such as sports cameras. For the retail market, we also introduced new consumer SSDs, such as the SanDisk Extreme II SSD, and initial reviews have been very positive.
Turning to SSDs. For the client market, we have completed several customer qualifications of our SSDs made with our 19-nanometer technology and we began revenue shipments of these products during the quarter.
We also began shipments of high-performance 19-nanometer Client SSDs in small form factors using the PCIe interface. The Client SSD market continues to grow, driven by increasing customer preference for high-performance, instant on capabilities, antenna and sleek form factors.
SanDisk is particularly well positioned to benefit from this trend as we are providing our broad array of SanDisk alone -- standalone and caching solution to the vast majority of leading PC OEMs already. In Enterprise SSDs, we are pleased to report another quarter of solid double-digit sequential growth, representing our third consecutive quarter of record revenue.
Further building on this momentum, we are excited with our recently announced agreement to acquire SMART Storage Systems, an innovative supplier of SATA and SAS SSDs for the enterprise space. This acquisition will help us further deepen our enterprise product offerings by building upon our successful line of at SAS SSDs.
Upon the close of the acquisition, which is expected in the third quarter, SMART Storage portfolio of SATA products will immediately expand our competitive capabilities and open up a $1.6 billion incremental opportunity for SanDisk in enterprise SATA. Additionally, the acquisition will also add 2 new Tier 1 storage OEMs to our customer roster and we will be qualified with our Enterprise SSDs as 6 of the top 7 storage OEMs upon close of the acquisition.
Further, this acquisition will enable us to augment our engineering team with talented engineers from SMART Storage Systems. SMART Guardian Technology, which is a suite of key firmware and hardware design technologies that together enhance the endurance and reliability of the Enterprise SSDs, has already won strong customer endorsements, and we expect to broaden its benefits by implementing it in our other future product offering.
As industry demand for Enterprise SSDs continues to grow, we believe that vertically integrated manufacturers, with captive non-flash supply, have a distinct long-term advantage in terms of both the development capability and supply continuity over others that must procure NAND. From a manufacturing standpoint, in the second quarter, our 19-nanometer technology represented approximately 85% of fab output.
We expect to begin initial production of 1Y technology late in the third quarter of 2013 with meaningful output commencing in the fourth quarter. Switching to supply topics, our forecast for SanDisk captive bit supply growth remains unchanged in the 20% range for 2013 and between 25% and 35% for 2014.
For the industry, our expectation for healthy supply and demand balance in 2013 and 2014 remains unchanged. Our bit growth estimates for the industry have consistently included the assumption of new wafer capacity additions, as well as conversions of some DRAM capacity to NAND and early generation of 3D NAND, and we have not heard anything new that would change our estimate.
As we mentioned at our Investor Day, we do not expect the early generations of 3D NAND to be cost-effective versus 2D planar NAND nor to have any meaningful contribution to the industry bit supply until the 2015, 2016 time frame. In late June, we announced our decision to commence construction of Phase 2 of Fab 5 with expected completion of the building shell in mid-2014.
I will remind you that the primary use of the new clean room space will be to enable the 1Y and 1Z technology transition in Fab 3, 4 and Phase 1 of Fab 5, and for a pilot line for BiCS NAND. We do not expect Phase 2 of Fab 5 to contribute to any meaningful increment to our total wafer output in 2014.
In closing, we are very pleased to have reported record second quarter revenues and strong margin, setting the stage for a stronger second half of 2013. With robust long-term demand drivers, coupled with our broadening portfolio of solutions and growing traction with our diverse customer base, we are excited about our future prospects.
We remain focused on executing to our technology strategy, prudently managing our capacity, driving profitable growth and creating shareholder value. With that, I will turn the call over to Judy for her financial review and outlook.
Judy Bruner
Thank you, Sanjay. We delivered robust second quarter results, reflecting the benefit of our diversified product portfolio and customer base, our low-cost leading edge memory supply and our disciplined capacity management.
And we are pleased to have accelerated returns to our stockholders with $280 million spent on share repurchase during the second quarter and $370 million year-to-date. Our record second quarter revenue grew 10% sequentially and 43% year-over-year.
Our ASP per gigabyte was up 5% from Q1, increasing sequentially for the third consecutive quarter, driven by a higher revenue mix of embedded products, coupled with a stable supply-demand environment. And this is the first time in our history that our ASP per gigabyte increased on a year-over-year basis, rising 6% from Q2 of last year, reflecting an increased mixed of SSDs and embedded products, along with a strong pricing environment.
The year-over-year shift in our product mix toward SSDs and embedded products is also manifesting itself in an increased average capacity of per unit sales. Average capacity per unit increased a strong 92% year-over-year.
Our bit sold increased 4% sequentially and 35% year-over-year, with the growth in our bit supply coming primarily from continued transition to 19-nanometer memory. Our second quarter revenue came 65% from our commercial customers and 35% from retail with strong year-over-year growth in both channels.
Our retail sales grew 22% year-over-year with continued improving mix of high-performance products and particularly strong growth in the average capacity of our retail mobile cards. Geographically, we achieved strong year-over-year retail growth in all major regions.
Our commercial revenue grew 57% year-over-year with embedded products and SSDs being the biggest contributors. Looking at the mix of our total revenue by form factor, the combination of embedded and SSD products achieved the highest ever proportion of our revenue at 46% with embedded revenue at 30% and SSDs at 16%.
Our removable product revenue was flat year-over-year with retail card and USB revenue growing enough to offset the decline from OEM bundled card sales and our de-prioritization of non-branded card sales. Our non-GAAP gross margin increased over 600 basis points sequentially to 46.7% in Q2, exceeding our previous forecast primarily because of better-than-expected pricing and product mix.
Our cost-per-gigabyte declined 5% sequentially and 21% year-over-year with the primary drivers being increased 19-nanometer mix and a weaker yen, partially offset by the impact of increased revenue mix from products containing a higher proportion of non-flash memory cost. The sequential improvement in our 19-nanometer memory mix was the result of beginning-to-sell Client SSDs using 19-nanometer technology.
The cost improvement due to the yen reflected a Q2 yen rate in cost of sales of 85 compared to 81 in Q1 and 79 in the year-ago quarter. Our Q2 non-GAAP operating expenses were $260 million, equal to our previous forecast.
Our Q2 non-GAAP operating margin was a robust 29% and our non-GAAP tax rate was 31.5%, slightly below the rate we had forecasted. Our diluted share count increased slightly sequentially as our accelerated share repurchases were only partially counted in weighted shares for Q2.
And this decrease was offset by the impact of our 2017 convertible debt, which added 1.6 million shares to diluted share count. For Q2 reporting, the 2017 convertible was in the money with a conversion price of $52.37 compared to the Q2 average stock price of $57.05.
Accounting rules do not allow us to count the offsetting shares expected from our call option, which will act to increase the effective conversion price of the convertible to $73.33. We have posted a schedule on our website, which outlines the share dilution expected from the convertible at various stock prices and also shows the expected offset from the call option at maturity of the convertible.
Turning to the balance sheet and cash flow, we generated our highest ever Q2 cash flow from operations at $391 million. We spent $928 million to pay off the remainder of our 2013 convertible debt and we spent $280 million on share repurchases to retire 5.07 million shares at an average price of $55.27.
Other key cash flow items included $71 million of expenditures for non-fab CapEx and a receipt of $20 million from the fab joint ventures. Our total gross cash and investments ended the quarter at $5.35 billion.
Our accounts receivable increased sequentially due to higher sales and due to the timing of both our retail and commercial sales being more weighted toward the second half of the quarter due to retail seasonality and OEM customer needs. The fab joint ventures invested in approximately $130 million of fab equipment for SanDisk technology transitions in the second quarter, bringing year-to-date fab investments to approximately $180 million.
Our off-balance sheet fab equipment lease guarantees stood at $612 million at the end of the quarter. I'll now turn to forward-looking commentary.
The forecast I will provide include the estimated non-GAAP impact of the SMART Storage Systems acquisition, which we expect will close before the end of our third quarter. We are seeing strong demand for all key product lines, and we expect supply-demand balance to remain healthy.
Our bit supply growth will be limited and will come primarily from continued transition to 19-nanometer memory, which is now being used broadly across our Client SSD products. Our third quarter revenue forecast is $1,525,000,000 to $1,575,000,000.
And we now expect our full year revenue to be between $5.95 billion and $6.05 billion, a substantial increase from our previous estimate. We forecast our third quarter non-GAAP gross margin to be in the range of 47% to 48%, reflecting a positive impact from a weaker yen exchange rate and a higher 19-nanometer mix, partially offset by some downward movement in our blended pricing, primarily due to product mix, including a seasonal increase in USB sales for back-to-school shopping and the mix of custom embedded products.
We now expect our full year gross margin to be north of 45%, higher than our previous forecast of 42% to 44%. We expect our expenses to increase in the second half of 2013 due to the SMART Storage Systems acquisition, the timing of certain R&D project expenses and increased headcount for development of our expanded product roadmap and to support our growing commercial customer base.
We forecast Q3 non-GAAP expenses to be between $285 million and $290 million and we now expect the full year non-GAAP operating expenses to be approximately $1.1 billion. We expect our tax rate to remain about the same, and we expect our diluted share count to be approximately constant with a continued strong rate of share repurchase, offset by increased accounting dilution from our convertible debt.
We expect 2013 total capital investments, which include investments from our flash ventures to be near the lower end of the range previously provided or approximately $1.1 billion with the total net cash outlay of approximately $350 million. In summary, we are pleased to raise our 2013 forecast for revenue and gross margins.
And we now expect to deliver a record year in top and bottom line results. We will now open the call for your questions.
Operator
[Operator Instructions] We'll hear first from John Pitzer with Credit Suisse.
Michael Chen
This is Michael Chen speaking in for John Pitzer. I just have a question on -- some of your competitors have stated that samples of 3D NAND could come out as early as end of the year.
Your expectations and time line for the production of 3D NAND is quite a bit later. So I was just wondering, I guess, can you please explain the difference or the discrepancy?
Sanjay Mehrotra
As we have described at our Investor Day, we are very much focused on continuing to scale NAND from 1Y, which we will begin to ramp up in production late this quarter to 1Z technology node, which we will begin to ramp up into production late next year, early 2015 time frame. And these are technologies that will be, we believe, will be the lowest cost technologies in 2014 and 2015 time frame.
Our 3D NAND technology that we plan to bring to production, the bit NAND technology is a highly scalable and very efficient technology, we will bring it to production in the marketplace, we believe, in second half of 2015, 2016 time frame because that's when this technology will give us lower cost than 1Z technology. So our 2D NAND and 3D BiCS technology continues to position us well for cost leadership through 2D NAND through 2015 time frame.
And then with 3D NAND, beginning in 2016 time frame. And also I want to point out that the 3D NAND technology that is being discussed by NAND industry supplier, is a technology that is not as scalable, not as efficient as our BiCS 3D NAND technology.
So from what has been published in the literature, we have studied that technology and we have concluded that it will take 2 or 3 generations of that technology before it becomes really competitive with 2D NAND technologies that will be in the marketplace. So our strategy is very much focused on maximizing the ROI of the technologies and that's why the longer you can go with 2D NAND, the better ROI you have.
And then once 2D NAND cannot be scaled any further, that's when it is the best to bring 3D NAND technologies. I feel very comfortable with our technology strategy for the years ahead in terms of cost leadership and continuing to drive our objective of profitable growth.
Operator
And our next question comes from Mehdi Hosseini of SIG.
Mehdi Hosseini - Susquehanna Financial Group, LLLP, Research Division
Sanjay, as you scale the SSD strategy on the enterprise side with all the investments, how should we think about the software piece of the revenue? And would you be able, at some point, to actually break out the software from the rest of the revenue mix?
Sanjay Mehrotra
The software revenue in the foreseeable future will continue to be small. And we do not plan to break it out.
Software really is giving us tremendous opportunity to engage in meaningful dialogue with enterprise customers for their future roadmap. And software, of course, is giving us opportunities to bring value, greater value, to our customers with innovative, differentiated solutions.
And our software is something that not only works with our hardware, our enterprise-grade hardware drive, but it is also a software strategy that will actually work with other competitors' hard drive as well, and that really, truly, brings strong value proposition to our enterprise customers.
Mehdi Hosseini - Susquehanna Financial Group, LLLP, Research Division
So how can we evaluate or quantify this value creation that the software piece brings?
Sanjay Mehrotra
Well, the software piece will give us value creation in our hardware solution sales that we make to the customer. In terms of expanding our opportunity and differentiating our value offering, as well as through software, we are able to expand the market opportunity for enterprise-grade SSDs as well.
Because the combination of hardware and software really provides even greater value and accelerate the penetration of SSDs in the enterprise segment.
Mehdi Hosseini - Susquehanna Financial Group, LLLP, Research Division
Sure, I understand. The problem that we're facing is the bits times ASP is no longer going to work for SanDisk, given the value creation that you just mentioned.
And I'm just trying to figure out if there is a better way of capturing and quantifying this value creation.
Judy Bruner
I think you'll see the value creation in our revenue growth and our margins over time.
Operator
And our next question comes from Hans Mosesmann with Raymond James.
Brian C. Peterson - Raymond James & Associates, Inc., Research Division
This is Brian Peterson in for Hans. Could you guys talk about what hedging assumptions you're using for the yen in the third quarter and then maybe in the fourth quarter?
Judy Bruner
Sure. Relative to the third quarter, we expect that the rate will be approximately 94.
And that is largely -- that is basically locked in, in terms of our purchases at this point. However, the exact rate will depend on which inventory flows through during the quarter.
And relative to the fourth quarter, our purchases that will impact the fourth quarter are largely locked in, although not totally locked in, and that rate at present, we expect will be about 97.
Operator
And our next question comes from Doug Freedman with RBC Capital Markets.
Earl Hege - RBC Capital Markets, LLC, Research Division
This is Earl Hege calling in for Doug Freedman. How should we think about cost declines moving through the year?
Just some of the puts and takes on that would be helpful.
Judy Bruner
Well, remember that our -- at our Investor Day, we indicated that we expect the cost decline for this year to be between 15% and 25%. And in fact, in the quarter just reported, the year-over-year cost decline was 21%.
And so we still expect to be in that range. I would expect that the yen will have a somewhat bigger impact on the third quarter than it did on the second quarter.
On the other hand, the increasing mix of 19-nanometer had a bigger cost improvement impact on the second quarter than we expect it will have on the third quarter. So those are some of the puts and takes.
Earl Hege - RBC Capital Markets, LLC, Research Division
Great. And one more, if I can.
How do you see the linearity of the 1Y ramp? In other words, when you -- when would you expect crossover?
And should we continue to expect an extended tail on the life cycle of that?
Sanjay Mehrotra
So 1Y, in terms of crossover in production output versus 19-nanometer, you would expect it sometime next year, sometime around mid-next year. 1Y will start production late this year, it will continue to ramp for several quarters.
I will expect it to ramp into 2015 perhaps as well. Remember, 19-nanometer started production late fourth quarter of 2011.
And has ramped through 2013 time frame. We just reported that in Q2, at least 85% of the total output.
I would expect a similar ramp in 1Y, perhaps, could be a bit longer because the mix of our embedded and enterprise business and OEM business now is larger compared to what it was during the 19-nanometer NAND.
Operator
And that comes from Mark Newman with Sanford Bernstein.
Mark C. Newman - Sanford C. Bernstein & Co., LLC., Research Division
My first question -- I have a question on your mix. I wanted to see, how much of this ASP increase is from the product mix versus increase in like-for-like, if there is any way you could break that out?
And then secondly, I have a question, to get a better idea of shortages in the industry or supply tightness, could you give us some kind of idea? Do you have customers that are actually on allocation?
Are you turning down business, which is helping you to improve your mix, because I -- we can see that your mix is improving quite substantially. Your mix in the removable has gone down substantially below 50% now.
And I'm wondering, is this being helped by shortages in the industry? And if you could give us some kind of idea in how severe the shortages are, to just get an idea what the situation is out there in the industry.
Judy Bruner
Okay, Mark, let me take that. So in terms of our price, our increase in price, absolutely it's been impacted by both real price increases that we have made in like-for-like products but it is also being influenced year-over-year and sequentially by product mix.
As our mix increases in embedded product, for example, which was very strong in the second quarter and year-over-year, very much -- there's a big increase in mix of SSDs. So both of those factors are increasing and increasing the average selling price.
Mark C. Newman - Sanford C. Bernstein & Co., LLC., Research Division
Is there any way to break out what the like-for-like increase would be Q-on-Q?
Judy Bruner
Quarter-on-quarter basis, it is more a like-for-like price increase than it is a mix effect on the average selling price. And then in terms on shortages and tightness in the market, we do absolutely have some pockets of supply constraint in our inventory.
And we are prioritizing carefully how we utilize our supply. And we are prioritizing supply for our value-added products.
We have reduced the mix in our business private-label cards, as well as wafers and components. And so that mix is also impacting our average selling price statistics.
Operator
And we'll move next to James Schneider of Goldman Sachs.
James Schneider - Goldman Sachs Group Inc., Research Division
Starting out, the industry commentary on bit growth. As I think you mentioned many of your -- the equipment suppliers, Semicon West last week talked about expectations of around 100,000 worth of new wafer starts equipment orders being done by the end of this year.
I was wondering, you certainly talked about the fact that you don't expect your bit growth estimates for the industry to fall outside the ranges, you outlined at the Analyst Day, 30% to 40%. But can you give us any sense about whether that wafer level do you believe is accurate and whether you think that might be towards the higher or lower end of that range for '14?
Sanjay Mehrotra
Yes, so for 2014, our estimates are that the total industry wafer production output will increase by about 10%. In 2013, that increases about 6% to 7%.
So if you look at 2013, we estimate that the total wafer production output in the industry would be around 14.5 million wafers. And in 2014 time frame, I would expect that to be somewhere under 16 million wafers, so amounting to about a 10% increase.
With that, I would just like to point out that, these are the lowest percentage increases in the industry in the last several years. I mean, the increases that we have had in 2012, toward 2013 and '14.
So this ultimately translates our estimations of bit growth in 2014 to about 30% to 40%. It is too soon to say whether it will be toward the upper half of that range or around the lower half but it will be somewhere in that range, we believe.
And with that 30% to 40% bit output growth in 2014, when you compare that to the demand increase that is expected, the underlying demand trend in the industry, I believe, it continues to bode well for a strong industry environment in 2014 as well. And these kinds of the bit growth rates, as we had highlighted at the Investor Day, 30% to 40% kind in 2013, as well as in 2014, these are the lowest ever bit growth rate in our industry.
And again, pointing to a healthy environment. So the supply trends, I believe, are best ever.
The demand trends are best ever in the industry. And overall, demand supply environment is healthy as ever.
James Schneider - Goldman Sachs Group Inc., Research Division
And then just a quick housekeeping one for me. Specifically, as we get to the Q4 of this year, what do you expect that 1Y node to constitute in terms of your total output or production, however you want to put it?
Sanjay Mehrotra
It will be -- in terms bit production, over 10%.
Operator
We'll hear next from Vijay Rakesh of Sterne Agee.
Vijay R. Rakesh - Sterne Agee & Leach Inc., Research Division
I just wanted to ask on the 1Y, 1Z transition, if you could lay out how do you think the mix would be as you ramp 1Y by the end of the year? Or how do you see 1Y next year?
Sanjay Mehrotra
1Y will continue to ramp during the course of next year, as I mentioned earlier. It will start in the third quarter, it will continue to ramp for several quarters, similar ramp, perhaps a little bit longer than the 19-nanometer ramp in terms of number of quarters.
But it will be ramping throughout the course of 2014.
Vijay R. Rakesh - Sterne Agee & Leach Inc., Research Division
Got it. And on the SSD side, obviously, you bought SMART Storage, any update on do you think your SSD mix will be much higher as you get through 2014 with the new acquisitions?
And do you see any, I mean, I know you have good visibility into the channel, but any of this can -- double ordering or any of that?
Sanjay Mehrotra
We are very excited about the pending SMART acquisition and this will definitely enable us to offer to our customers a broader array of offerings in SAS products, as well as really opening up this $1.6 billion enterprise SATA SSD market for us. And as I mentioned, I mean, a great team of engineers that gets added to the SanDisk already great team of engineers as well.
So this is really very exciting, and we are looking forward to bringing strong value proposition to our customers with the full benefits of vertical integration. And we remain on track for 25% of our revenue coming from SSDs in 2014 time frame.
And 2013, I would expect our revenue from SSD to be more than 15%.
Operator
And we'll hear next from Daniel Amir of Lazard.
Daniel L. Amir - Lazard Capital Markets LLC, Research Division
Given that you had an Analyst Day in the middle of the quarter, what was kind of the big difference here in the past, in the last 6 weeks of the quarter, that made the quarter much better than you expected? I mean, we are was really the upside, was it pricing, was it more of the product mix?
Was it the costs that were maybe slightly better? Because given that you clearly outperformed here on a big side on the margin front especially, and on the guidance here for the year, raising the revenue -- the annual revenue guidance.
Judy Bruner
Sure, Daniel. The biggest impact that enabled us to achieve higher revenue and higher margin than we had forecasted for the second quarter was the pricing that we experienced.
So it was really the very strong supply-demand environment and our ability to capture higher prices for the sales that we generated, that was the biggest factor. I would tell you that the costs were about as we had expected and it really was higher pricing.
Strong demand, of course, but higher pricing. And then a secondary factor was that even within certain product lines, we had favorable mix within those product lines.
So as an example, within our Enterprise SSDs, we had a favorable mix of products that impacted gross margins favorably.
Daniel L. Amir - Lazard Capital Markets LLC, Research Division
Okay. And then the follow-up question is, Sanjay, you gave some commentary about smartphones and card slots and the fact that microSD is doing well.
A few quarters ago, the industry was talking about a slowdown of cards and attach rates with cards and phones. Has anything changed in terms of that trend, given that more and more are moving towards more embedded?
Or was that more of a few quarter issue and you still believe that there is a market -- a large market here for the removable card side as well?
Sanjay Mehrotra
So what we had talked about last year was about bundling of cards with the mobile OEMs. And that trend has continued showing a decline in bundling rates.
In fact, bundling of cards with headset shipments is quite low and contributes a very small portion of our revenue at this point. So really no change in terms of the bundling of cards with the mobile handsets by the OEMs.
In terms of the opportunity -- some of that opportunity clearly with less bundling has shifted to the retail market and SanDisk, with its strong global share in retail, #1 brand, has really done well to capture that. And particularly that trend happens to be one of higher capacity cards as well, the smartphones, increasing appeal of those smartphones, increasing multimedia functionality with card slots.
When consumers buy the cards, they tend to buy highest capacity cards than what would've been in the past bundled with the phone. So it really is playing to our strength of a very strong global retail presence, as well the highest capacity of the retail card and certainly, high-performance as well because these smartphones are requiring more and more high-performance experience through the cards and our Ultra and Extreme cards, microSD cards, are doing well in the marketplace and we command a premium as well.
Operator
And moving on, we'll hear from Michael Bressler of Kensico.
Michael Bressler
I have 2 specific questions. My first question is on the NAND market.
The NAND market consists of both low-grade and high-grade products and the demand for high-grade product is increasing just as it is becoming more difficult to produce. How should I think about the price premium for NAND that is high-grade and can qualify for an enterprise drive?
And what percentage of your NAND that could qualify for an enterprise drive or this high-end application? And then how do you compare that with the overall market?
And I have one more question.
Sanjay Mehrotra
So with respect to the flash that is used in enterprise-grade products, certainly, that is high-quality flash, and SanDisk produces high-quality flash and with the benefits that we have with vertical integration and we have always said that in the enterprise space, it's very important to have the full benefits of vertical integration to really ultimately win in that marketplace. As part of the benefit of vertical integration, when the system engineers, the controller engineers are really able to work with the device technologies that are developing the flash, the interaction really develops in helping us produce at the wafer level, high-grade, high-quality flash, which we believe, is better than other suppliers in the industry.
And I'll just point to you that the flash memory that is used in our enterprise-grade product is the X -- 2-bit per cell, the X2 memory, not the X3 memory. So one aspect is leveraging vertical integration to really produce high quality flash memory wafers in production that ultimately can meet the requirements of enterprise grade products.
The second aspect, of course, is having controllers and the system expertise and the firmware to really work with the flash, to manage the flash, to really give the reliability and the performance that the enterprise-grade applications expect. And it's not only that controllers, system expertise and high-grade flash memory wafer production, but it is also the testing of flash memory at the back-end.
We really apply sophisticated testing techniques to really ensure that this flash memory is the highest grade flash to be used in enterprise-grade applications. So all of this really points to the benefits of vertical integration in leveraging high-grade, high-quality flash that is needed for leadership in enterprise grade.
Regarding the price premium in enterprise, of course, we have mentioned before that we have higher-than-corporate averages, the highest margins in enterprise-grade products. But we're really not prepared to disclose the price premium of enterprise-grade products versus other products.
Michael Bressler
Got it, okay. And my second question is on 3D NAND.
When I was at Semicon West, I met with a leading inspection company and they told me that the problem with 3D NAND is that you don't inspect the product until it is finished. So if the defect is discovered in the middle of the stack, there is no way to correct for it.
In the absence of great controller technology, how can this product ever be cost-competitive against leading-edge planar X3?
Sanjay Mehrotra
So there are 2 things here that I think you are referring to, Michael. One is, with the 3D NAND technology, in terms of the device structure itself, which you can look at it as multistory, multilayers of NAND flash stacked higher, certainly, in terms of in-line inspection or failure analysis, the challenge is much more daunting compared to the 2D NAND because, again, if there is a fault occurring at any level in a multi-level structure, it becomes -- it takes longer to figure that out as well as the failure analysis becomes more complex and you may have processed those wafers in the fab through all the steps until the end before you figure out that something has gone wrong at certain levels embedded deep in that multilayer structure.
So this points to the challenge of in-line inspection and failure analysis. And this ultimately, I believe, will translate to a longer yield ramp of 3D memory, as well as longer time before 3D memories become applicable across a broad range of applications.
So that's certainly a challenge of 3D memory, I believe, they will take longer time before they really become worthy of meaningful, cost-effective production due to the challenges of in-line inspection and failure analysis and manufacturing aspects. But certainly on the other hand, 3D NAND memories will have new failure mechanisms, which will certainly require system expertise, controller expertise to really productize them and make them meaningful volume production technologies.
And this is where SanDisk for 25 years has really, from day 1, focused on system expertise, all of the system expertise that we have, the lead system expertise, I believe, best in the world system expertise that we have here for our 2D NAND memory, will be applicable to our 3D NAND memory, but it will then require even more system expertise because 3D NAND will have its own unique considerations. And again, SanDisk's experience in system expertise will be able to extend, I believe, more readily in 3D memory.
So you are correct that in-line inspection is a challenge on wafer manufacturing side. And you are also correct that controller expertise is going to be a key to make 3D into a viable production-worthy technology.
Operator
And that comes from Ambrish Srivastava with BMO.
Gabriel Ho
This is Gabriel for Ambrish Srivastava. I have a quick question regarding the MCP.
So we understand the MCP is one of the fastest grower in Q2 and is used in the lower-cost smartphone, which is the fastest grower within the smartphone segment, so does that put you in a disadvantage since you did not manufacture your captive DRAM supply?
Judy Bruner
As we've said a number of times, we do earn a lower-than-average gross margin percentage on our MCP products because we source the mobile DRAM on a non-captive basis. But we do still earn very good overall gross margin dollars from our MCP products.
Of course, they have a higher cost, but they also have a higher price than our average. And the real value of the MCP for us is in the iNAND.
These are all MCP iNAND products. And so we really are getting paid for the iNAND in these products.
And keep in mind, we had a fairly significant increase in our MCP iNAND sales in the second quarter and still posted a very big increase in our gross margin.
Gabriel Ho
I see. As a follow up, could you also we remind us, what is your bit -- what is your wafer growth assumption for the 2014, given your range of 25% to 35% bit growth?
Judy Bruner
Sure. I had indicated at our Investor Day that for this year, for 2013, that with productivity improvements we expect to achieve about 2.6 million wafers for the year.
And that for next year, for 2014, our estimate is between 2.6 million and 2.8 million wafers.
Operator
And our next question comes from Craig Ellis of B. Riley.
Craig A. Ellis - B. Riley Caris, Research Division
Judy, just a follow-up on the CapEx comment, looks like there is a minor decrease in the planned CapEx for the year. What's that attributable to?
Judy Bruner
Nothing real significant. It's just getting more granular on our estimates and the timing of various tech transition spending.
Craig A. Ellis - B. Riley Caris, Research Division
Okay. And then switching gears on the downstream inventory side.
Can you just describe how you currently perceive downstream inventories? And related to an earlier question in environments like this where there is such a high degree of tightness in the market, we often see double ordering, how does the team ensure that that's not impacting shipments that head out the door?
Judy Bruner
Sure. We get very frequent reporting of our retail channel inventory and we believe we have good visibility of the distributor channel inventory related to our commercial channels.
And we don't see anything unusual in any of those areas. The channel inventory remains healthy and relatively lean in both our retail channel and the OEM channels that we do utilize.
Operator
And we'll hear from Steven Fox with Cross Research.
Steven Bryant Fox - Cross Research LLC
Just one question, just tying together some of the comments around sales timings and account -- versus account receivable and some of the comments you made about what changed since the Analyst Day. The strength in the second half of the quarter, I guess, sounds like there is retail seasonality, but breaking it down further, I was curious if there is anything you would comment on relative to the general health of the demand in your core markets versus, say, just product cycles and sort of ASP momentum?
And then if you could carry that through for the rest of the year, if you're seeing anything outside the ordinary of seasonality in your new guidance?
Judy Bruner
Sure. I don't think that there were any issues in terms of health of the demand and the timing of the revenue that we were -- that we experienced in the second quarter.
As you indicated, you're right that the sell-through timing of the retail business tends to be more in the second half of the second quarter and we did experience that. And then relative to our commercial channels, the revenue was also more weighted to the second half.
But it appears to just have been relative to the needs of our various OEM customers, their product launches, their product timing and I do not see anything unusual. Our accounts receivable, by the way, is very high quality and very current accounts receivable.
I don't think there is any indication of demand issue. Relative to the rest of the year, I haven't seen anything that would cause me to think that the timing or seasonality of our revenue would be different than historical.
Operator
And the next question comes from Harlan Sur of JPMorgan.
Harlan Sur - JP Morgan Chase & Co, Research Division
Your average gigabit per unit shipped almost doubled on a year-over-year basis for the second consecutive quarter. Was most of that driven by a higher mix of embedded or did you see average capacity going up across most of your segments and do you think you can sustain this type of capacity growth going forward?
Judy Bruner
Both are true. We saw good average, very strong, average capacity growth really across all of the key product segments that we sell.
But in addition, of course, mix also played a factor. As year-over-year, the embedded mix of our business has grown significantly and that tends to be higher average capacity than some of the removable products.
And -- but in particular, SSDs, grew significantly in mix and those of course, have the highest average capacity. So we are seeing a good mix effect in that average capacity.
But we are seeing very strong growth in the really all of the different product lines. I mean, including imaging and USBs and even commercial card type products.
Harlan Sur - JP Morgan Chase & Co, Research Division
Great. And then on your client SSD business, does the team expect that to contribute to quarter-on-quarter growth in Q3?
And you've got the partnership with Western Digital targeting hybrid drives. How does the team view the adoption of hybrid drives versus full-blown SSDs as we move through the second half?
And given the rollout of these new Gen 3 ultrabooks?
Sanjay Mehrotra
We certainly do expect quarter-over-quarter increases in our SSD revenue from Q2 to Q3 and we expect SSD revenue to be increasing percentage of our business in the second half. And regarding the hybrid, hybrid opportunity really is just beginning.
It is too early. 2013, we do not expect much contribution from hybrid drive.
Flash -- flash sold by us -- flash solutions sold by us for hybrid drive. But we certainly expect that to increase in 2014 time frame.
And as new Haswell platforms are launched and sold in the channel in second half of this year, that certainly bodes well for Client SSD drives because, Haswell platforms will give a longer battery life, and that should be a nice appeal to drive growth of Client SSD and ultra-thin notebook category.
Operator
And moving on, we'll hear from Steven Chin of UBS.
Stephen Chin - UBS Investment Bank, Research Division
Two quick ones for me. First one is on your mobile embedded business.
I was wondering if you could provide some color on what the average capacity might be for, I guess, your iNAND versus your iNAND MCP [ph] products and I was wondering if you have any visibility on demand content going to lower-cost smartphones versus premium smartphones?
Sanjay Mehrotra
In MCP, the average capacity of flash tends to be lower. MCPs tend to be used in volume more in the budget smartphone or mid-range as well.
And discrete iNAND tends to be more in the midrange and high-end smartphones. So that tends to have higher average capacity.
And if you -- we don't break it out separately between, I mean, we draw that color here, but if you look at the total embedded average capacity, it is close to our corporate average.
Stephen Chin - UBS Investment Bank, Research Division
Got it. And my second question is regarding [indiscernible] drives and other enterprise-related flash products.
Just with the SMART Storage deal you're ready to close sometime this quarter, also your Pliant acquisition a couple of years ago also running smoothly right now. Can you provide any color on your target for when overall enterprise sales might reach 5% or 10% of sales?
Sanjay Mehrotra
We are not really breaking it out between client and enterprise. As we mentioned that in 2014, we expect our clients plus Enterprise SSD sales to be 25% of our total revenue.
And we are making really great progress on both, on Client SSD, as well as Enterprise SSDs. Client SSDs are, in total, revenue larger than Enterprise SSD revenue but we have strong momentum in both of them.
Operator
And that question will come from Kevin Cassidy of Stifel, Nicolaus.
Dean Grumlose - Stifel, Nicolaus & Co., Inc., Research Division
This is Dean Grumlose calling in for Kevin. I was wondering if you could provide some color around the outlook for the company's long-term target model as you shift more towards embedded and SSD products?
Is this increase in margin less sustainable over time? And how may the corporate targets change as you progressed?
Judy Bruner
Sure, Dean. I don't know if you're aware, but I did just adjust and increase the long-term financial model at our recent Investor Day in May.
And I did increase the gross margin range of that model, as well as the OpEx range because it does require more OpEx to be investing in some of these product areas, in particular the solid-state solutions. And we are also investing heavily in our three-pronged memory technology strategy.
But net-net, I did increase the operating margin range by 100 basis points to 21% to 25%. And I said at the time that I thought there is long-term potential to further increase our long-term financial model as, in particular, SSDs and somewhat embedded but more SSDs and enterprise, in particular, becomes a larger portion of our model.
So I do think there is that longer-term potential to further adjust the model. But of course, at this point, we did just adjust it and it's too soon to talk about further adjustments to that model.
But I do think there is upside to gross margin and to operating margin as our mix further evolves.
Brendan Lahiff
Thank you, everyone. We want to thank everybody for joining our call today.
A webcast replay of today's call should be available on our Investor Relations website shortly. Thank you, and have a good evening.
Operator
Ladies and gentlemen, again, that does conclude today's conference. We do thank you all for joining us.