Apr 26, 2007
TRANSCRIPT SPONSOR
Executives
Lori Barker Padon - Investor Relations Eli Harari - Chairman of the Board, Chief Executive Officer Judy Bruner - Executive Vice President, Administration & Chief Financial Officer Sanjay Mehrotra - President, Chief Operating Officer
Analysts
Jim Covello - Goldman Sachs Craig Ellis - Citigroup Amit Kapur - Piper Jaffray Daniel Amir - WR Hambrecht+Co Satya Chillara - Pacific Growth Equities Sameer Doctor - J.P. Morgan Eric Gomberg - Thomas Weisel Partners
Operator
Good day, everyone and welcome to today’s first quarter 2007 earnings call. As a reminder, this call is being recorded.
For opening remarks and introductions, I would now like to turn the call over to Ms. Lori Barker Padon.
Please go ahead, Madam.
Lori Barker Padon
Thank you. Good afternoon and welcome to the financial teleconference for SanDisk Corporation for the first quarter of 2007.
I am Lori Barker Padon, SanDisk's senior investor relations director. Joining me is Dr.
Eli Harari, Chairman and CEO of SanDisk; Sanjay Mehrotra, President and COO; and Judy Bruner, Executive Vice President of Administration and CFO. The agenda for today’s teleconference is as follows: Eli will start with remarks about SanDisk and trends in our markets; Judy will follow up with our first quarter financial results and future guidance; and we will close then the teleconference with your questions for Eli and Judy.
Any non-GAAP financial measures 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.
After the completion of this call, an audio replay of this conference, a copy of today’s prepared comments and quarterly metrics will be made available on SanDisk's investor relations website at SanDisk.com. During our call today, we will be making forward-looking statements.
Any statement that refers to expectations, projections or other characterizations of future events or circumstances is a forward-looking statement, including those related to revenue, pricing, expenses, gross margin, tax rates, inventory, production capacity, technology transitions, joint ventures and future products. Actual results may differ materially from those expressed in these forward-looking statements.
Certain factors that may cause actual results to differ are detailed under the caption Risk Factors and elsewhere in the documents we file from time to time with the SEC, including our Form 10-K for fiscal 2006. We undertake no obligation to update these forward-looking statements which speak only as of the date hereof.
We do not intend to update information contained in this teleconference. Now, I would like to turn our call over to our CEO, Eli Harari.
TRANSCRIPT SPONSOR
Eli Harari
Good afternoon, everyone. I will discuss current market conditions and recent product announcements, update you on recent IP developments, provide you some interesting thoughts on the demand/supply situation, and then Judy will provide you our Q1 financial details and the Q2 business outlook with her remarks to follow.
The first quarter was a down quarter for SanDisk and for the industry as a whole. Excess supply of NAND MLC, coupled with soft seasonal demand in retail precipitated sharply lower pricing.
Our first quarter price reductions in retail, although quite steep, trailed the much sharper price reductions by our competitors and probably caused us a small loss of market share in U.S. retail.
Notwithstanding price increases recently announced by our several NAND component suppliers, we continue to see a substantial amount of low-priced inventory of cards and flashdrives from our competitors in the channel. Accordingly, we have taken pricing actions for the second quarter to narrow the pricing gap between us and our competitors.
We currently expect price reductions to moderate in the second half of the year, given the projected strong pickup in demand as we head into the holiday sales season. There were a number of bright spots in the quarter.
One was the record 27 million mobile units sold; indicative of our continuing penetration into what we believe will be our largest opportunity for flash storage in the coming decade. Another was the high level of design-in activities for our newest products, including iNAND, solid-state drives, and megaSim, as well as increasing adoption of our TrustedFlash security platform.
And of course our cross-license agreement with Hynix, which I will discuss a little later. This week Dell announced the immediate availability of the 32-gigabyte solid-state discs supplied by SanDisk for two of their corporate notebook models.
Dell has also disclosed their intention to propagate the deployment of solid-state discs in a substantial number of their future notebook platforms. We are seeing strong interest from leading notebook OEMs in our 1.8-inch and 2.5-inch solid state products.
We believe that solid state discs must transition from SLC, single-level cell, to MLC NAND to achieve pricing that is attractive to main-street consumers. We anticipate that this transition to MLC NAND will favor SanDisk over our competitors, because solid state discs, unlike flash cards or USB flash drives, impose demanding specifications that are not easily met with NAND multi-level cell.
SanDisk knows MLC like no other competitor. SanDisk, together with legacy M-Systems, has deep industry experience and related IP in flash solid state systems.
We expect to begin transitioning our solid state disc products to MLC later this year and we anticipate modest revenue in 2007 with acceleration in 2008. In mid-April, Sony and SanDisk announced our agreement to jointly specify, develop, and co-own a new super high-speed flash card format for professional video capture, using the PCI express bus for rapid downloads of large video files to a PC.
We believe that this new format, called the SxS, although initially a small niche may over time find other high performance storage applications outside of video. In early April we began shipping the Sansa Connect in partnership with Yahoo!
Music. The Connect has been receiving excellent product reviews, and is perceived by pundits as pushing the envelope in WiFi connectivity and music sharing.
On the lower end of our MP3 players, we began shipping last week the Sansa Shaker, a really fun and affordable MP3 player designed for kids and families. In the mobile space we began shipping 4-gigabyte micro SD, 4-gigabyte Memory Stick Micro, that’s the M2 cards, and 8-gigabyte iNAND, all industry firsts.
I will comment briefly on Hynix, STMicro and Qimonda. In late March we signed a long-term cross-license and supply agreement with Hynix.
This agreement avoids protracted, costly litigation and reaffirms the value of our patent portfolio. It is important also because it clears the way for future cooperation between SanDisk and Hynix, for example in the area of x4 technology.
With regards to STMicro, the Texas court has provided a 90-day stay in the trial involving patent infringement claims by both companies. The ITC case is awaiting the judge’s ruling, which is scheduled for June.
Yesterday, Qimonda and SanDisk announced an agreement to form a joint venture in the fast growing market for multi-chip packages for embedded storage in handsets. Qimonda is one of the leading specialists in low power mobile DRAM which is used extensively in many of these multi-chip packages.
Qimonda previously announced that it has de-emphasized its internal Flash NROM development activities. Going forward, Qimonda will use our NAND flash die and controllers in some of their multi-chip packages, while we will correspondingly use Qimonda’s mobile DRAM for some of our MCPs.
This will augment our embedded storage products for handset vendors and we expect to ship those first multi-chip packages products from this joint venture in the first half of 2008. I will now discuss our progress with our 56-nanometer technology transition and our NAND supply.
In the first quarter of 2007, along with our partner Toshiba, we began ramping up volume production of 56-nanometer NAND MLC at our 300-millimeter fab 3. In the second quarter, we expect to ship 56-nanometer products incorporating 16-gigabit and 8-gigabit chips.
Actual 56-nanometer product revenues recognized in the second quarter will be modest. However, they are expected to grow substantially in the third quarter and become the majority of bits shipped in the fourth quarter.
Particularly encouraging is the excellent progress we have been making on 56-nanometer yields, which have been improving faster than expected. We anticipate achieving cost per megabyte parity with the mature 70-nanometer NAND MLC technology sometime this quarter, ahead of plan.
This important milestone considerably reduces the risks associated with the 56-nanometer technology transition and bodes well for our cost reduction later this year. Wafer capacity at fab 3 is also ramping to a maximum capacity, which we now estimate will reach 150,000 wafers per month before the end of this year.
This is higher than the 135,000 wafers per month that we previously projected for fab 3, the increase coming from improved efficiencies at essentially no additional CapEx investment. By year-end, we expect our captive output to be overwhelmingly on 300-millimeter wafers processed on 56-nanometer technology.
Looking beyond 56-nanometer technology, our NAND MLC joint development team is targeting production starts of 43-nanometer wafers in the first half of 2008. Fab 4 is on plan for a September ribbon cutting ceremony, and it is expected to start contributing to our sales in the first half of 2008.
Given the current excess supply of NAND and NAND production cutbacks announced by some of our component competitors, it is fair to ask why we are not cutting back production from fab 3 or slowing down fab 4. I will address this question: First, we project that our demand drivers in the fourth quarter will likely require the full output from fab 3.
Consequently, given the faster-than-expected ramp of our 56-nanometer technology, we will likely procure less non-captive supply than previously expected for 2007. In addition, in 2008, we expect output from fab 4 to augment our supply needs to meet our projected demand.
Second, in times of fast growing demand from new markets, capacity can quickly flip from excess to shortage and result in customer allocations. There are a number of potential catalysts for driving demand in 2008, including the impact of the iPhone and other multimedia phones on consumption of 4-gigabyte to 16-gigabyte flash in mobile phones, the adoption of 32-gigabyte to 128-gigabyte solid state disks in notebook PCs, and the continuing proliferation of NAND flash into higher capacity music players, video Players, personal media players and gaming consoles and so on, not to mention external stimulus from events such as the Beijing Summer Olympics and the 2008 U.S.
election year. In this changing environment, guaranteed low-cost supply becomes a powerful tool to maximize profits and an important lever in expanding our retail and OEM market share in support of our new products, markets, and geographies.
Fab 3 is now achieving exceptionally competitive costs and as long as it is operated at the leading edge of technology and fully utilized, we believe that the return on our investment in Fab 3 will continue to be favorable. The third point, the NAND flash market may soon experience structural changes that could potentially bring about periods of considerable supply shortages over the next two to three years.
Specifically, we believe that the drastic price declines for NAND flash in 2006 and 2007 will have the unintended consequence of accelerating the obsolescence of industry-wide 200-millimeter NAND capacity. Currently it is estimated that more than 50% of the world’s supply of NAND flash is manufactured on 200-millimeter fabs.
That capacity will not be viable by the end of 2008, except for low chip densities, which are a relatively small part of the market. Replacing this capacity, let alone incrementing it, will require several new 300-millimeter dedicated NAND megafabs.
If not on the drawing board today, such a fab will not add to NAND capacity before 2009 at the earliest. Compounding this issue, most of the current 200-millimeter DRAM capacity will also become obsolete by the end of 2009 or sooner, when DRAM has to transition to 78-nanometer and below, where these fabs cannot operate.
The combined obsolescence of NAND and DRAM 200-millimeter capacities on such a large scale is unprecedented. This scenario in fact is the exact opposite of what the NAND industry experienced in the last two years, where ample 200-millimeter DRAM capacity was thrown into the NAND supply equation.
So I hope that this discussion helps you understand our steadfast resolve to execute our long-term manufacturing strategy. In order to maximize long-term value creation, we believe that the benefits of assured low-cost supply outweigh the pain of short-term imbalances, such as excess inventory that inevitably materialize during industry down-cycles.
We are keenly aware of our shareholders’ concerns for the short term; however we believe that we are executing the correct strategy for achieving SanDisk’s long-term strategic goals for market leadership in the growth years ahead. I will now turn the call over to Judy.
Judy Bruner
Thank you, Eli and good afternoon everyone. I’ll begin with comments on our revenue.
Our product revenue increased 28% year-over-year, resulting from an increase in megabytes sold of 209%, a decline in ASP per megabyte sold of 62%, and the consolidation of the USB joint venture between former M-Systems and Toshiba. On a sequential basis, our megabytes sold were down 22% and our ASP per megabyte was down 23%, with the ASP decline much more aggressive in the retail channel.
For the first quarter, the mix of our product revenue was 49% retail and 51% OEM. I’ll organize the majority of my revenue comments by retail sales and OEM sales.
Our retail sales exhibited more seasonality this Q1 as compared to last Q1. Our retail units sold were down 32% sequentially compared to down 19% sequentially in the prior year.
This exaggerated Q1 decline may be related to the particularly strong fourth quarter of 2006, in which we experienced 52% sequential unit growth compared to 37% growth in the previous Q4. Compounding the seasonally down unit sales was an aggressive retail pricing environment in which our retail ASP per megabyte was down 35% sequentially and relatively low growth in average capacity from Q4 to Q1.
On a year-over-year basis, retail units sold grew 50% and average retail capacity increased 87%, but with ASP per megabyte down 67% year-over-year, our retail revenue was down 8% year-over-year. Looking at our retail sales by end market, we experienced strong year-over-year unit growth in all five of our key end markets, with the strongest year-over-year unit growth in mobile, MP3 and gaming.
Our year-over-year unit growth in retail USB flashdrives was equal to our average retail unit growth while unit growth in the digital still camera market was below the blended average but still up 39% year over year. Analyzing our OEM revenue is more complex because of the addition of the former M-Systems business.
On a sequential basis our OEM unit sales grew 4% and ASP per megabyte was flat, but our OEM revenue declined 8% sequentially due to a lower average capacity. We are now selling embedded NAND products which are lower average capacity and higher ASP per megabyte than our legacy SanDisk products.
With a full quarter of this business in Q1, compared to six weeks in Q4, statistics on average capacity and change in ASP per megabyte are distorted. This is why I am providing more granularity this quarter on our retail versus OEM sales and pricing to help you understand our results.
For our legacy SanDisk OEM business from Q4 to Q1, unit sales were approximately flat and ASP per megabyte was down 9%, a much less aggressive price change than in our retail business, but this is largely timing related. On a year-over-year basis, our legacy SanDisk OEM business experienced unit sales growth of 80% and our consolidated OEM business had unit sales growth of 130%.
Consolidated OEM average capacity increased 69% year-over-year and ASP per megabyte decreased 54% year-over-year, again influenced by the addition of the embedded business. In total, our OEM revenue was up 109% year-over-year, which included growth in the SanDisk legacy OEM revenue of 25% year-over-year.
The strongest OEM growth came from the mobile market in which unit sales grew 117% year-over-year. Compared to last year, OEM revenue also benefited from the addition of the former M-Systems embedded industrial business and the white label USB drive business.
While we have significantly de-emphasized the white label USB business, it did still contribute to first quarter revenue, as did the consolidation of the USB joint venture with Toshiba, called TwinSys, which added $53 million to Q1 revenue. Our royalty revenue in the first quarter of $97 million was above our previous forecast, primarily due to the mix of component royalties from our licensees and also stronger card-based royalties.
Our Q1 non-GAAP product gross margin of 18.5% was down 14 percentage points sequentially and 10 percentage points year-over-year. Our ASP per megabyte declined faster than our cost per megabyte, especially in the retail channel where the sequential decline in unit sell-through in the first quarter meant that a high percentage of our cost of sales came from product manufactured in 2006.
Another factor in our Q1 gross margin was that inventory reserves taken in Q1 as compared to those taken in Q4 had an incremental impact on gross margin of approximately 4.4 percentage points. Our Q1 shipments were made with 5% non-captive memory, reflecting good progress in converting former M-Systems products to captive memory.
Non-GAAP operating expenses of $177 million declined $8 million from Q4 and this was after adding in a full quarter of the former M-Systems expenses and $6.5 million of charges for restructuring actions taken in the first quarter. Seasonal and volume-related declines in certain sales and marketing expenses contributed to the sequential reduction in expenses, as did the expense containment actions announced in February which took effect quickly and included a reduction in force, a freeze in salaries and a reduction in executive salaries, as well as other expense cutting measures.
Other income of $36 million increased $4.6 million sequentially, primarily due to higher cash balances. Our non-GAAP tax rate of 39.8% in Q1 reflects a forecasted annual non-GAAP rate of 35.2% plus $3.8 million of discrete items that impact Q1 only.
Fully diluted shares for Q1 reflect a full quarter of the shares issued for the M-Systems acquisition. Turning to the balance sheet, cash and investments increased $217 million to $3.5 billion.
First quarter cash-flow from operations was $255 million. Accounts receivable decreased very significantly, reflecting strong collections, reduced Q1 shipments and the accrual of significant reserves for promotions and price protection for inventory in the channel.
Inventory, net of reserves, increased $98 million to $594 million. At approximately 90 days of inventory, this is more than we would like to carry, but we do expect to utilize this inventory to expand our international footprint over the course of this year.
We invested $44 million in property and equipment, primarily for equipment and building construction at our captive back-end assembly and test factory in Shanghai. We received $25 million of repayment against our note receivable from the 200-millimeter FlashVision venture, the third such repayment since we stopped investing in the 200-millimeter fabs.
Fab 3 equipment investments in the first quarter were funded by operating lease draw-downs of $349 million. Our total net operating lease guarantee now stands at approximately $958 million.
I’ll now turn to our outlook for the second quarter. Given the typically stronger demand in the latter half of the second quarter, we expect significant sequential growth in both units and megabytes, partially offset by continued decline of the white label USB business.
On the pricing front, as Eli described, we have made price reductions in Q2 in order to bring our pricing more in line with our competitors. One other factor which will impact revenue, although not bottom line earnings, is that we have agreed with Toshiba to terminate the TwinSys USB joint venture formerly between M-Systems and Toshiba as of the end of Q1.
The consolidation of this joint venture contributed $53 million to our first quarter revenue. Taking into account the expected growth in units and average capacity, lower Q2 prices, the elimination of TwinSys and the decline in white-label USB sales, we expect Q2 product revenue to be in the range of $625 million to $725 million.
We are forecasting royalty revenue in the range of $95 million to $100 million for Q2. This reflects an estimated decline in our licensees’ Q1 NAND revenue, offset by new license revenue from our Hynix agreement which is being recognized on a straight-line basis over the life of the agreement.
We expect NAND royalty revenue to recover in Q3 based on the higher MLC component pricing we are now seeing in the market. For 2007, we currently expect license and royalty revenue of approximately $400 million.
Turning to gross margin, we expect price declines to be approximately offset by cost declines leading to a forecasted non-GAAP product gross margin for Q2 of 15% to 20%. For the full year 2007, we continue to believe that the non-GAAP product gross margin will be in the 15% to 25% range that we provided at our February analyst day.
We expect Q2 GAAP product gross margin to be lower than non-GAAP by approximately 3 points due to the inclusion of stock compensation and acquisition-related purchase accounting charges. We forecast Q2 non-GAAP operating expenses to be in the range of $190 million to $200 million reflecting approximately $15 million of fab 4 R&D start-up costs and some increase in sales and marketing expense associated with Q2 seasonal programs and costs related to expanding our international footprint.
Q2 GAAP operating expenses are expected to be higher than non-GAAP by approximately $37 million due to stock compensation and amortization of acquisition related intangible assets. Other income for Q2 is expected to be between $30 million and $32 million, as cash is forecasted to come down due to fab investments and lower collections in Q2.
The minority interest on our P&L will be zero with the termination of the TwinSys joint venture, and diluted shares for Q2 are expected to be approximately flat, with share repurchases, which began during Q1, offsetting estimated new issuances. On the balance sheet, we forecast that inventory will increase in Q2.
We expect this inventory to be absorbed by growth in demand in the second half of the year, including expanded distribution in various international geographies. We expect to make cash investments in fab 3 equipment in Q2 of approximately $275 million.
While our outlook for this year remains consistent with the model we presented at our February Analyst Day, we are clearly not satisfied and across the company, we are focused on driving innovation, market share gains, cost reduction and investment in the areas strategic to our future. The NAND supply/demand balance remains an issue for the industry and SanDisk.
However, we are optimistic that balance will return in the fourth quarter of this year and beyond as the growth of new and existing markets is fueled by lower price points, and as supply growth is challenged by the future requirement for 300-millimeter fab capacity, which Eli described. When NAND supply/demand balance returns, we expect an extended period of much more modest price declines, translating into a more robust financial model in 2008 and beyond.
We will now open the call for your questions.
Operator
(Operator Instructions) We will take our first question from Jim Covello of Goldman Sachs.
Jim Covello - Goldman Sachs
Good afternoon, guys. Thanks so much for taking the question.
A few questions. I guess first, could you help us understand maybe a little more of the detail around the Hynix announcement?
I know you talked about what the full-year royalty revenues would be but I think there is a fair amount of confusion in the marketplace about what exactly the technologies that Hynix agreed to pay licenses on. If you could give us some kind of color of the royalty rates, vis-à-vis what some of the other licensees might be paying, that would be great.
Eli Harari
As you know, these agreements are all confidential. All we can say is that it is a long-term agreement the involves a cross-license for a NAND component, including SLC and MLC, as well as a supply agreement to SanDisk -- that is, Hynix will supply SanDisk at preferential terms.
We have also said that we will work together on putting together a joint venture entity that will develop jointly, focused very much on x4, four bits per cell development and manufacturing for the two companies. Other than that, we really cannot disclose anything else.
Jim Covello - Goldman Sachs
But it definitely does include SLC and MLC?
Eli Harari
Yes.
Jim Covello - Goldman Sachs
Okay. Another issue I just wanted to touch on, relative to --
Eli Harari
Let me just -- I’m sorry, you know, the market is gone, is essentially now MLC, almost entirely, so any license that excludes MLC would not be very attractive to anybody.
Jim Covello - Goldman Sachs
Sure. I think there was some speculation in the marketplace after the original announcement that it was only by four, but thanks for helping to clarify that.
A question I guess on the potential supply tightness in the second-half of the year. There have been some that have speculated that some of the NAND capacity that got moved over to DRAM is going to come back into the NAND market before the second-half of the year, preventing supply from getting tight.
I wondered if you had a view on that.
Eli Harari
You have heard the comments from our Korean suppliers who are the main drivers of capacity, as well as of course Toshiba and SanDisk. We are very focused on, and I just gave you what our plan for capacity increases in fab 3 as well as adding more megabytes through the 56-nanometer.
I think that the pricing that we saw in first quarter very, very much in my opinion was a consequence of pricing MLC. Basically, a lot of MLC came to the market in fourth quarter from companies that were trying to move from SLC to MLC for quite some time and finally made it happen in the fourth quarter.
The market was unable to absorb that MLC, in my opinion, either because of performance issues or because of controllers not being able to handle it. As a result, MLC was priced at -- you know, if you look at the Gardner and other, you know, this is all public information.
MLC went from being priced about 10% below SLC to being priced at half of SLC, and that created the whole market for SLC as well as MLC because most of the sales were in MLC. What you have seen recently is corrections, significant corrections from some of these suppliers to bring the price of MLC much closer to that of SLC, within about 10%.
We think that that is very healthy for the industry. Of course, it is very good for us from the point of view of royalties on MLC that were being depressed when the MLC price went down by a half.
It has now gone up very close to SLC. SLC I believe -- SLC and MLC are priced approximately correctly, in my opinion, relative to the cost structure that we see.
For us, we have always said a vast majority of our output is MLC, probably 98%, 99%, and therefore we have just a single price, and of course we do not sell a component, we sell an end product. So it is not only in the -- so I think that what -- we are very focused on demand/supply in balance but frankly, the good news here is that we think the pricing for MLC in the industry has been corrected and the only problem that we have now is that there was some damage done.
There was a lot of inventory that was sold in the market at very low prices of MLC. Until that inventory clears the channels, we still -- that is the price in the market for cards and flash drives.
Once that is flushed out, I think that we will see a much more moderate pricing environment.
Jim Covello - Goldman Sachs
Terrific. One more last one and then I promise I will go away; just in terms of the non-captive supply.
Judy, I think you said it was only 5% this quarter?
Judy Bruner
Correct.
Jim Covello - Goldman Sachs
And then you would expect it to remain roughly at that low level in the next couple of quarters? I am just trying to gauge the impact of the higher margin captive supply on the model going forward.
Judy Bruner
Currently we expect the non-captive to remain at low levels.
Jim Covello - Goldman Sachs
Terrific. Thanks so much.
Lori Barker Padon
Next question.
Operator
Your next question comes from Craig Ellis of Citigroup.
Craig Ellis - Citigroup
Thanks. Good afternoon, everybody.
Cycling back on the inventory issue, just to understand how that dynamic is playing out, are you saying that in retail or at OEM that inventories have peaked and they are starting to work down, or are they still building a bit? How is that playing out in both sides of your business?
Judy Bruner
Craig, I would tell you that channel inventory is at a very comfortable level. We ended the first quarter with channel inventory of about nine weeks, which is measured based on backward-looking sell-through, so that is a good level.
In fact, it is a little less in terms of weeks than we had at the end of Q1 of last year. So where the inventory is a little heavier than we would like is the inventory on our balance sheet.
Craig Ellis - Citigroup
Okay, and then just understanding the pricing dynamic that is related to that, are you talking about pricing in the second quarter that looks like last year and the prior year in the second quarter? Or are you talking about something that would be much greater than seasonal?
Judy Bruner
Well, we have not given a specific guidance on a pricing range for the second quarter but we would be hopeful that the pricing is less in Q2 than it was in the first quarter in the retail environment, but we have moved a number of prices lower in all geographies this quarter, as Eli said, to try to close the gap with our competitors in retail.
Eli Harari
Craig, basically what we said is that price reductions that were projected for 2007 over the whole year are back-end loaded. The first-half of the year is taking the brunt of that and we think that once that takes place, it will slow down very significantly because frankly, the pay level will be very, very difficult for all the suppliers, even the best local supplier.
Craig Ellis - Citigroup
Is it just broad elasticity off of that that’s giving you the confidence to raise your output for the year for fab 3, or is it a particular product program that you see, whether it be the MP3 products or cards or something else?
Eli Harari
As you know, if you look at the last six years, we have a very clear trend heading into the fourth quarter, and we have given some guidance for the year of our expected growth in terms of megabytes, so we do -- we are very enthusiastic and very optimistic about the demand for our output in the fourth quarter.
Craig Ellis - Citigroup
Lastly, if I may, Judy, this is the first time I have heard you call out a card contribution on the royalty line. Can you give us some sense of some of the relative contribution that cards your new royalty has and the traditional royalty payer as we exit the year, and maybe something about how those different levers can grow over time?
Judy Bruner
The increase in the card-based royalty that I referred to was a portion, explained a portion of the increase in royalty revenue that we had over the guidance that we had previously given. The majority of our royalty revenue still remains based upon the sale of NAND flash components.
In terms of the new royalty payer, as I mentioned, this is actually license revenue that is impacting us today, which is being amortized on a straight-line basis over the life of the agreement. As I said, the increase from that license revenue approximately offsets the decrease that we expect to receive this quarter based on component-based royalties, which as you can estimate, would come down given the very significant price decline in Q1 in the NAND component pricing, which impacts our royalties in Q2.
Clearly for confidentiality reasons, I cannot give you any specifics, but hopefully those comments help you try to estimate some of these numbers.
Eli Harari
But let me add on the card license, this is not from any specific one license fee. Most of the increase in card license revenue comes from the success of SD, and most importantly micro-SD, SD, mini-SD, micro-SD are now growing at a very fast pace.
We, together with Matsushita and Toshiba, collect a royalty on every one of these cards that are sold, so our share of that has grown.
Operator
Anything further, Mr. Ellis?
Craig Ellis - Citigroup
That’s it. Thank you for taking the questions.
Operator
Thank you. Next we will here from Amit Kapur of Piper Jaffray.
Amit Kapur - Piper Jaffray
Thank you very much. I was wondering if you could update us on the expansion of your international retail distribution points and maybe talk about some of the key geographies that you are targeting.
Sanjay Mehrotra
In terms of expansion on the international, we are very much focused on expanding the distribution network as well as having more feet on the ground in Europe and Asia. In that regard, after the M-Systems acquisition, as part of our integration we have really leveraged the M-Systems team to help us manage and grow our retail sales footprint in Europe and Asia.
We have just put all that in place in the first quarter. We are making good progress there and we would expect increased share gains and increasing retail storefront presence in Europe and Asia in the future.
Amit Kapur - Piper Jaffray
Maybe turning to the mobile handset market, when you talk with the handset OEMs, how do you see the mobile market playing out? Do you think it is still primarily a memory card opportunity, or are you starting to see embedded flash ramp pretty aggressively in devices?
Sanjay Mehrotra
Actually, we are seeing embedded flash ramp nicely but the largest part of portion we believe in the future remains on the card side. But as you know, on the embedded side, we have strong solutions with iNAND as well as the acquisition of M-Systems, NDog, and continuing to have high-capacity products.
Now, with this week’s announcement of Qimonda, our ability to improve our offerings on the MCP side we believe will give us a very good leading position on the embedded as well as card solutions. And you see a good mix of opportunities on both sides.
Amit Kapur - Piper Jaffray
Thank you very much.
Operator
We will take our next question from Daniel Amir from WR Hambrecht.
Daniel Amir - WR Hambrecht+Co
Thanks a lot, a couple of questions here. First of all, it seems like the demand supply environment in the second-half of the year, maybe into early ’08 is impacted by probably two major market, the solid state drive market and the video market as well.
Can you comment a bit about how SanDisk views the market, how you think that plays out for you guys here in the next 12 to 18 months?
Eli Harari
We think that the second-half of this year as well as the next year is still mostly driven by the handset market. That is the biggest opportunity.
We have said in fact the first quarter was the first quarter that mobile revenues exceeded card revenues for digital cameras, which is a very significant milestone and we expect that to grow. The biggest opportunity is -- catalyst are much more like iPhone than solid state disks.
Not to take anything away from solid state disks, but as I have indicated we see that as really a 2008 significant contributor. In the second-half of this year, it is still going to be iPhone, iPhone like devices from their competitors that required 4- and 8-gigabytes of storage, as well as audio players, MP3 players that go through this stage -- and video will be there but relatively still small because the content for video is not yet what -- the infrastructure for video delivery in small devices and handset devices is still going to take some time.
But we do see, once you get -- even on the card side once you get cameras that can take the HDSD -- that is the SD cards with the 4-gigabyte and higher capacity, which is basically the entire market is moving to that, then we are going to see the card business also move past this 2-gigabyte capacity and go to 4- and 8-gigabyte. So it is a very broad spectrum of markets and applications, not any one of those, but definitely the handset market is where we see our very strong position, both on the OEM and on the retail, both on the embedded and the removable.
In the future, also the MegaSIM.
Daniel Amir - WR Hambrecht+Co
The second question is in your comments you prepared that you are planning to do some retail price cuts here, or you started doing them as well. Is the strategy here basically to come back and gain market share after losing market share in Q1?
If that is the strategy, should we see that going into the second-half of the year, the company is largely more focused on market share rather than necessarily maintaining a certain gross margin level?
Eli Harari
Market share is very important to us and to a large extent -- frankly, in this particular case, we are not driving pricing in the market. We are really following the consequences of these very steep 50% price declines that were not initiated by us.
Frankly, we were trying to hold back and we got [pegged] for that and we are not happy with that. We are always trying to balance market share with profitability, maximizing profitability.
At the end of the day it is all about profitability, but 2007 is a tough year. There is no denying that.
We believe we do have today the best cost structure out there, that we do have the ability to go and take market share and yes, we will do so.
Daniel Amir - WR Hambrecht+Co
Final question is where do we stand on the x4 development now falling Hynix agreement? When should we expect production and what applications are you focusing it on?
Eli Harari
x4 has been under development by legacy M-System for the last more than five years. I cannot get into details of how it is being developed but there has been ongoing development that we believe is on technology that is not leading edge, and basically we are -- it is very similar to what we have done frankly with the 3D technology that was running on, you know, two generations behind and we are moving them as quickly as possible to leading edge.
We are doing the same thing with x4. That means that the initial expectations for product maybe later this year will not happen, simply because even with x4 with four bits per cell on an older generation technology competing with two bits per cell on the generation that we expect to be in volume production in the second-half of this year will just not be competitive.
So this is really I would say probably the second-half of 2008 type of product introduction for x4 on a much more leading edge technology.
Daniel Amir - WR Hambrecht+Co
Thanks a lot.
Operator
Moving on, we will here from Satya Chillara of Pacific Growth Equities.
Satya Chillara - Pacific Growth Equities
Good afternoon. Eli, this question is for you regarding the Hynix agreement.
I wanted to dig a little more, see if you can answer any of them. How long is this agreement for?
Hynix did NAND production of $5 billion so far cumulative revenues. Did you guys receive any catch-up payments?
Lastly, if you can talk about is it at least similar to what you are getting from your other licensee? That would be appreciated.
Eli Harari
Of course, we cannot discuss that, as you well expect. I think Hynix will be very upset and everybody else that is listening there, our competitors would also be wanting to know -- anyway, I can’t comment.
Satya Chillara - Pacific Growth Equities
On all three questions, or -- on the agreement, how long is the agreement? At least can you talk about that?
Judy Bruner
Even the length of the agreement is confidential.
Satya Chillara - Pacific Growth Equities
Okay, got it.
Eli Harari
But it is a long-term agreement.
Satya Chillara - Pacific Growth Equities
Okay, understood. So in terms of, Judy, your guidance seems to be pretty soft compared to what Eli was calling on the TV today, 625 to 725 product.
Can you give us the assumptions, what you are thinking here in terms of bit growth and pricing?
Judy Bruner
What I can tell you is that it does reflect significant growth in units and bits in our ongoing business, but remember that that is partially offset by the continued decline of the white label USB business, as well as the elimination of the consolidation of Twinsys, and I pointed out that Twinsys added $53 million of revenue in the first quarter. So there definitely is what we believe is good growth in units and bits, offset by the two items I just mentioned, and also offset by the pricing decline in the second quarter.
Satya Chillara - Pacific Growth Equities
Lastly, Eli for you on x3, three bits per cell. How long have you been working on this program and what is the confidence level at this point going into -- before you start the production, do you feel pretty comfortable on x3?
What is the timeframe we are talking about here?
Eli Harari
x3, as I said on February 26th, is going to be very important technology. Our target there is to introduce x3 with no compromise in performance, whereas x4 is much more limited in scope of applicability.
We think that x3 has got to be basically just a lower cost version of x2 or MLC. We are making good progress on that.
There are some joint development activities that are going on and we are on plan to introduce the first product. I am not sure that we -- okay, so we said beginning of 2008, so I think that is -- let’s put it like this; in early 2008 is when we hope to introduce the first x3 product.
Satya Chillara - Pacific Growth Equities
Okay, but how long are you putting the R&D? Can you answer that, please?
How long have you been working on this?
Sanjay Mehrotra
For over -- between one to two years.
Satya Chillara - Pacific Growth Equities
Great. Thank you.
Eli Harari
Let me say that everything that you need to know in x3 you need to develop first in x2. You need to have very fast x2 in order to be able to deliver a cost-effective but high-performance x3, so you could say that we have been developing the foundation for x3 for probably over the last 10 years.
Lori Barker Padon
Thank you. We have time for two more callers.
Operator
Thank you. Our next question comes from Sam Doctor of J.P.
Morgan.
Sameer Doctor - J.P. Morgan
Thank you. I have a couple of questions.
First of all, on the OEM pricing, how should we see that in the second quarter? Is there some kind of a catch-up on the ASP declines in the second quarter, given that your first quarter declines were lower than you expected?
Judy Bruner
You know, let me say a couple of things on the pricing front. We are not going to provide specific price guidance, in part because this is competitive data in the marketplace.
But over time, the net price of our products between retail and OEM are not that different in terms of the net price but there can be pretty significant timing differences from quarter to quarter in the different products. So we have taken into account the overall blended price decline in our revenue guidance for next quarter.
Sameer Doctor - J.P. Morgan
When you are talking about low-capacity NAND for Vista or Robson PCs or hybrid hard disks, is a 200-millimeter fab competitive for those kinds of applications? That would be mostly --
Eli Harari
When I say low-capacity, I mean more 1-gigabit, 2-gigabits type of chips that are used in multi-chip packages. Vista, Robson and these guys are going to go to a gigabyte and that takes an 8-gigabit -- it is very difficult to compete against an 8-gigabit chip with four 2-gigabit chips.
It is impossible to compete with a 16-gigabit chip -- against a 16-gigabit chip cost structure with four 4-gigabit chips, even if you are running them on a totally depreciated fab.
Sameer Doctor - J.P. Morgan
But for hybrid hard disks that have seen mostly you have 256-megabyte and 512-megabyte as the standard right now.
Eli Harari
Right, and if you calculate, even if they are very successful, if you calculate the amount of flash it is actually not very large. These fabs, these mega fabs really require applications that are driving 4-, 8-, and 16-gigabytes type of applications.
This is really what we are very, very focused on.
Sameer Doctor - J.P. Morgan
Thank you.
Operator
Thank you. We will take our final question from Eric Gomberg from Thomas Weisel Partners.
Eric Gomberg - Thomas Weisel Partners
Thanks. A couple of questions.
One, I was just wondering on the receivables, that is as low as I have ever seen relative to the revenue. Just wondering if you could just briefly comment on what is going on there.
Do OEMs pay slower?
Judy Bruner
Sure, Eric. So there are some anomalies this quarter in terms of the receivables and it partly has to do with the fact that, as you know, we recognize our revenue for the retail business based upon sell-through, so there can be anomalies when, for example, we had high shipments and thus receivables at the end of Q4 and that gets collected in Q1.
But we are recognizing revenue in Q1 based upon sell-through and accruing price protection and promotions for that sell-through, which can often be inventory that was sold and already collected from the customers. So you should not look at the DSO as being a meaningful statistic this quarter.
Eric Gomberg - Thomas Weisel Partners
Okay, fair enough. A question on Hynix, realizing you are very limited in what you can say.
You said royalty is straight-lined over the life of the agreement. I am just wondering; as far as x4 or any development there, is that included in the payments?
Or, should that be developed successfully, is that potentially incremental? And then just wondering if you could talk about expectations on additional licensees, royalties from others or renewals over the course of ’07.
Eli Harari
x4 is separate from the current cross-license agreement and the supply agreement. It is a separate -- it is an MOU, memorandum of understanding that we have with Hynix to explore joint development and manufacturing joint venture for x4.
So this is actually a work in process but it is not -- no, it does not -- it is not part of a payment that we are receiving for the cross-license. It is something that is going to stand on its own two feet and is a good business proposition for both Hynix and SanDisk.
Your second question?
Eric Gomberg - Thomas Weisel Partners
That’s very helpful; expectations on additional royalties or renewals from others over the course of ’07.
Eli Harari
Well, on the NAND side, on the NAND supply side, we have today by our estimate somewhere between 90% and 95% of the NAND supply is licensed. Really, other than Micron, there is no other major supplier out there.
I had indicated on February 26th that we will be more aggressive on enforcing our IP on the system side, system and card side and we see there is an upside over there, but on the NAND side it is pretty well covered with the exception that I said.
Eric Gomberg - Thomas Weisel Partners
Fair enough, and just one last question; is it fair to assume that the white label, which is trending lower and the Twinsys is lower than corporate average gross margin, and also just in terms of thinking about cost structure with 300-millimeter fab 3 ramping as well as it is, would we expect that internal costs are declining at a faster pace in the second-half than in the first-half?
Judy Bruner
In terms of gross margin, I would say yes, the white label business has a lower gross margin than the corporate average. But that is different than Twinsys.
Twinsys, I think you should think about that, as that comes out, it is neutral to the bottom line because remember revenue, gross margin comes out but also this minority interest, which is negative currently to the bottom line. That comes out as well.
I’m sorry, your second question?
Eric Gomberg - Thomas Weisel Partners
It was just that given the trajectory of the ramp on 300-millimeter and with the 8- and 16-gig monolithic at 56-nanometer, should we assume that internal costs are accelerating, that cost declines are accelerating in the back-half of the year versus what you are experiencing in the first-half?
Judy Bruner
Yes, the second-half of the year will definitely have more benefit from the 56-nanometer ramp than does the first-half of the year. Remember, as you know there is always a lag based upon the fact that we recognize a significant portion of our revenue based on sell-through, which can often be one to two quarters after we ship it.
Eric Gomberg - Thomas Weisel Partners
Okay, great. Thanks a lot.
Eli Harari
Thanks, Eric. I think it is time to wrap it up.
I really want to thank everyone here for listening today. We are definitely in a challenging year, 2007 we have said it will be a challenging year, first and second quarter, but we are very enthusiastic and quite optimistic about improving conditions later on in the year.
We are very focused on getting our costs and our innovation engine totally focused on end of this year and 2008. We are looking forward to seeing a number of you this quarter during investment conferences.
Thank you very much.
Operator
This concludes today’s call. We thank you for your participation and have a wonderful day.
TRANSCRIPT SPONSOR