邁威爾科技 (MRVL) 2007 Q3 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Marvell Technology Group Ltd. third quarter fiscal year 2007 financial results conference call. My name is Letitia, and I will be your coordinator for today. [OPERATOR INSTRUCTIONS]. For opening remarks and introductions, I will now turn the call over to Dr. Sehat Sutardja, Chairman and CEO of Marvell Technology Group. Please go ahead, Dr. Sutardja.

Thank you, Letitia. Welcome, everyone, to our third quarter fiscal year 2007 conference call. Weili Dai, Chief Operating Officer, and George Hervey, Vice President of Finance and Chief Financial Officer, are joining me on this call.

Today, we reported Q3 revenue increased 22% from the prior year to $520 million. Sequentially, our Q3 revenues decreased 9% from the prior quarter, which is slightly better than our revised guidance of a 10% decline that we issued on October 7th. Although we are disappointed with our sequential decline in revenue this last quarter, we remain excited about our continued strong positioning for growth as we expand our presence into many new large volume markets. In particular, we are very excited to have now closed our purchase of Intel's application and communications processor business earlier this month. With this acquisition, we are pleased to have this world-class engineering team join Marvell.

By combining these world-class products with our own technology, we are now strongly positioned to be a leading supplier in the cell phone market. With our broadening IP portfolio, Marvell has clearly become a very diversified and full fledged semiconductor company. With the breadth of our IP, we are now in a position to serve substantially all the large semiconductor markets. If we are not currently shipping into such a market, we are internally developing the solutions to enter them. Our ability to successfully enter such a broad number of semiconductor markets is due to the strength of our core technology, and our ability to repackage our technology and apply it to the specific end market requirement.

I will elaborate more about our opportunities and our business progress, but first I will have George provide more insight into our Q3 financial results.

Thank you, Sehat. Good afternoon, ladies and gentlemen. I would like to remind all participants that the following dialogue will contain predictions, estimates, and other forward-looking statements. Current subjects such as enterprise, consumer and emerging market trends, competition, customers, suppliers, products and demand, and expectations regarding revenue growth, gross margin expectations, operating expenses, other income, accounts receivable and inventory.

Such statements will be preceded by the words like expects, anticipates, believes, should, will, may, or words with similar import. These statements include those relating to our guidance for the fourth quarter of fiscal 2007.,the pace of our business for fiscal year 2007 and the impact of the acquisition of the XScale business of Intel on our operating results. The following factors, among others, could cause results to differ materially from those described in the forward-looking statements.

They include the inability to further identify, develop and achieve success for new products, services and technologies, the inability to successfully integrate the application, and communication processor acquired from Intel, increased competition and its effect on pricing, spending, third-party relationships and revenues, as well as the inability to establish and maintain relationships with commerce, advertising, marketing and technology providers. We direct your attention to our annual report on Form 10-K and quarterly reports on Form 10-Q, recent current reports on Forms 8-K and other Securities and Exchange Commission filings, all of which discuss other important risk factors that may affect our business, results of operations and financial condition.

Please be reminded that we undertake no obligation to revise or update publicly any forward-looking statements for any reason.

Now, moving to the Q3 '07 financials. As we indicated in the press release, we will only be reporting revenue for Q3 fiscal '07. Today we reported our net revenue for the third quarter of fiscal 2007 was $520.4 million, an increase of 22% over the $426 million reported for the comparable quarter of fiscal 2006 and a sequential decrease of 9.3%, from the second quarter of fiscal 2007. The sequential decline of 9.3% from Q2 fiscal '07 is slightly better than our October revised guidance of a decline of 10%.

As we indicated in our October realized outlook, the decline in net revenue was largely due to lower than expected demand from a number of our hard disk drive customers. The lower demand was caused by a combination of weaker than normal seasonal shipments in the personal computer market, as well as excess component inventory held by some of our large storage customers. Our networking products contributed another solid quarter as we continued to gain market share. In the embedded market our products continued to increase demand and we increased our shipments to a new gaming platform. We continued to experience lower demand for [radio] 2.11 end products as that product continues to be negatively affected by an excess of product in the retail channel.

For Q3 our enterprise markets represented approximately 80% of revenue with consumer the balance. We had three 10% customers in Q3, Western Digital, Samsung and Toshiba. While we cannot provide detailed numbers for our gross margin and operating expenses, the following are trends that we experienced during the quarter. As we indicated in our Q2 conference call, we expected a less favorable product mix for our revenue in Q3 resulting in a slight decline in gross margin percentage from Q2. Based on our October revised outlook, Q3 product mix and volume-related absorption had more of an unfavorable impact on gross margin percentage than originally planned. For the quarter, our gross margin percentage generated on shipments, the market pricing environment, and manufacturing performance were consistent with our revised expectations.

For operating expenses, our headcount additions and fixed spending items were as planned. As we indicated in our October update, we expected to incur substantial costs related to the ongoing internal review of the Company's historical stock options practices by the Special Committee. Actual Q3 costs for the review were significantly higher than expected. Our balance sheet continues to remain very strong. Our Q3 cash increased to $876 million, with $798 million in Q2.

Our inventory trends were as follows. Due to the lower than originally planned shipments for Q3, the quantity of product in our inventory increased. We have taken steps to bring our days of inventory down to support our ongoing business, and we expect to reach that point in Q1 fiscal '08.

Now I would like to turn the call back to Sehat for comments on our business outlook.

Thanks, George.

Last week, we announced the completion of our acquisition of the application and communication process of business from Intel. We are fortunate to have this tremendous opportunity for such a talented group of employees, proven world-class technology and strong customers join the Marvell family. Strategic opportunities to enter such a large volume market are rare. To truly capitalize on this acquisition, and to successfully integrate this business, we have a tremendous amount of work and investment ahead of us. One of the main tasks at hand is for us to quickly port the existing as well as new designs over to our foundry partners to help drive down product costs.

For a number of different reasons, the de-cost structure of these products is currently too high. As we started the business, it is apparent that most of the factors that have led to this high product cost are centered around manufacturing-related issues, such as the products being built in low volume, captive manufacturing facilities, that utilize 8-inch technology. Also a variety of other manufacturing decisions were made with an overall focus on product performance, at a sacrifice of cost.

Now, the resulting technology and performance of these solutions are the best in the market nothing else comes close. This is clearly evident by the large market share garnered by these solutions in the high end SmartPhone market.

We are confident in our ability to reduce the costs as we migrate the production to [advanced] 65-nanometer geometries and implement our aggressive design rules. By leveraging our own mixed signal IP and advanced libraries along with Intel's low power circuit design techniques, which by the way the industry's lowest power design methodology we will drive a considerable amount of costs out of these products, which will enable us to achieve [gains] consistent with this market. Also by driving down the product cost, we will enable our customers to continue and leverage the investment in these solutions and introduce forms into the higher volume mainstream SmartPhone segments.

At the same time, we will continue to aggressively invest in this technology to integrate more features and further enhance performance, which will enable our customers to maintain their leadership position in the high end, full featured SmartPhone segment. Although this is no small task, we are confident in our ability to make this a successful and a relatively quick transition. What is very important to note is that this business is based much more than just the semiconductor products but also on the complete platform solutions, software, design, and verification tools as well as the years of extensive qualification work that has been made. There's been thousands of man-years of investment placed into these solutions by both Intel, as well as their customers.

While the transition of the products to our Atlas model will help lower costs, our customers will be able to continue to leverage their own investment that they have placed in these solutions. We are very excited to be supporting such world-class customers as REM, Motorola, and (indiscernible).

The feedback we have received from our customers is very strong, given Marvell's proven track record as a supplier capable of integrating very complex technology and bring such solutions cost-effectively to the mass markets ahead of the competition. Also, our customers are excited about the breadth of our solutions for the future of the cell phones, including our Wi-Fi, Bluetooth and multimedia capabilities. Based on the strength of our solutions and the size of the market, we see tremendous growth opportunities ahead. In order to capitalize on these opportunities, we will need to continue and build upon the significant level of ongoing investment into this business.

The resources required to be balanced, advanced 3G and next generation technology is considerable. Although the current level of operating expense for this business is high, relative to the current revenue level, we believe this level of investment is required to be a serious supplier in this market even with this tremendous level of investment required up front, we believe that this is the correct entry into this exciting market. After all, if we wanted to get into this business from scratch, it would have also cost us a tremendous amount, but would have taken much longer time and we would have missed the important inflection point with SmartPhones and 3G technologies beginning to ramp.

By acquiring this business, we are getting into the cell phone market at an exciting point in time, and with very strong technology, as evidenced by the world-class designs, these solutions are shipping into today.

To date, this business has been capacity-constrained which has limited its ability to fully capitalize on the tremendous technology that has been developed. As we transition the production to our fabless model and remove many of its constraints, we believe we will be able to drive the revenue growth required to quickly rationalize our ongoing level of the investments being placed into this business.

Finally, I would like to highlight the strength of the existing technology and its competitive positioning in the marketplace. Our cellular-based technology has worldwide carrier acceptance and offers the highest throughput and proven 3G solutions which have passed I/OT testing by multiple customers. The XScale operation processors are also in a league of its own. The comparison to competitive solutions in the market, such as TI's OMAP processor is straightforward. On any given process geometry, the XScale processor is at least twice as fast as the fastest OMAP.

Another way of saying it, the OMAP is equivalent to two process generations ahead behind -- I mean, behind the XScale. Also, if you run our processor at the same operating speed, the XScale consumes about half the power.

The reason XScale has such a significant performance advantage is that OMAP uses a standard arm processor. However, the XScale processor, just like our own Ferocian processor is an internally developed processor, happens to run the arm instruction set. Intel has invested heavily for over ten years with world-class designers to develop the high speed, low active, and standby power technology implemented in these processors. These technology is so fundamental that any company that wants to achieve such power and performance metrics will have to copy this part of the technique. Our customers clearly see the benefit of such technology, as evidenced by the significant investment placed in designing their solutions around these XScale technologies.

Now by combining our original microprocessor development efforts with the new XScale team, we look to even further improve the performance, power and features for our customers.

On a personal note, I would like to emphasize that I have dedicated a great deal of my time to resolve all of these issues with this business. I have also worked closely with the key managers of this business and we are confident that together we can solve these problems. This will require a great deal of time and focus for the next year and I'm putting my personal reputation on the line that we will succeed. This is a tremendous opportunity and with our success, it will truly take Marvell to the next level.

Now, moving to our existing businesses. First, we continue to be very excited about the continued adoption of embedded wireless LAN in a growing number of applications. The adoption rate continues to grow as wireless LAN increases its penetration into a number of high volume applications, such as cell phones, gaming, cameras, printers, and personnel media players. We continue to lead this trend with our low-power Wi-Fi and maintain our substantial market share lead in the United States. We expect this only to accelerate, given the strong design activity around our recently introduced single chip wireless LAN and Bluetooth device.

For 802.11n, the retail market is still faced with inventory challenges which we believe will persist through the balance of Q4. However, we are very active with our retail customers and continue to be strongly positioned with top designs across all the major suppliers which will position us well when the existing inventory levels are worked off. I want to stress that the performance and the reliability of our 802.11n solutions are enterprise class and as a result, we are also winning very nice designs for deployment in leading enterprise equipment.

Additionally, for the same reasons, we are doing very well for the enterprise-class designs, we are also experiencing design success for the emerging home video applications. We have [formed] designs with leading OEMs for such equipment, including major TVs, as our solutions offer the highest reliability video distributions for the home.

Now, moving to gigabit ethernet. We continue to enjoy solid revenue growth as a result of our strong competitive positioning with our total solution of software, silicone, and [RADLAN] designs. In the infrastructure market, we continue to do well across all segments and we see particular strength in the SMB and Soho markets, given our smart managed line of solutions.

In the hard disk drive market, we saw a decline in our revenues for the first time this last quarter as we faced some near term inventory issues with some of our customers. Despite the current revenue shortfall, we remain very focused on continuing to develop to expand our technological lead in the marketplace. No other company is matching the level of investment that we are placing into our storage solutions. With the tremendous investments in our technology, we have continued to move even further ahead of the competition, and provide our customers with the highest level of performance and cost of ownership in the industry.

Finally, we are making good progress with our optical storage solutions of late with our development of overall system level solutions. However, we underestimated the sheer number of modes of operations that need to be supported in today's DVD market. As a result, it has taken considerably longer than we expected to develop all the software required for truly complete solutions. However, we are addressing the software issues by placing more resources on the software development, both internally and with third-party partners.

The good news is that recent customer feedback has been very encouraging, as they acknowledge the benefits, strength and reliability of our DSP-based solution. We look forward to giving more updates as our customers get closer to volume production.

I would now like to turn the call back to George for additional comments regarding our financials and guidance for the next quarter.

Thank you, Sehat.

As mentioned in our Q2 call, we would be updating our quarterly guidance once we had completed the acquisition from Intel of their XScale application processor and communications processor business. We completed that acquisition on November 8th and the guidance we will be providing on this call will include the XScale business.

Before moving to the guidance, I would also want to provide an update on the other parts of Marvell's business. As I mentioned in my Q3 commentary, due to excess component inventory held by some of our large [merged] customers, we experienced lower demand for our storage products in Q3. As we enter Q4, we believe that much of the inventory issue has been resolved and we expect customer demand to modestly improve in Q4. Market conditions remain stable for our communications products, such as networking and printer ASIC products, but overall demand for embedded wireless LAN products with the gaming market will be lower than Q3. Based on any of those factors we believe that our Q4 revenue from existing products should increase to approximately $530 million.

In addition, the partial quarterly revenue from the XScale business [four] should contribute approximately $90 to $100 million, resulting in total Marvell revenue increasing approximately 20% from Q3. Our gross margin percentage and operating expense percentage will also be affected by the acquisition of the XScale business, however, we expect gross margin percentage on our existing business to improve slightly from Q3, and operating expenses to increase slightly more than the rate of revenue growth for our existing business. As we discussed in our conference call, when we announced the plan to acquire Intel's XScale communication, and communication processor business, the business has been operating at a loss and would be dilutive to our consolidated financials in Q4 and improving going forward.

We are currently estimating the dilutive impact of Q4 of $0.05 per share with a reduction in overall corporate gross margin percentage and increase to operating expenses. In order to provide adequate working capital for the Company, including to fund the purchase of Intel's application and communication processor business, we successfully secured a three-year term debt financing in the amount of $400 million. This debt will have the effect of reducing interest income, the effect of reducing EPS by $0.02 per quarter.

Now I would like to turn the call back over to Sehat.

Thanks, George. That completes our commentary. Letitia, would you please poll for questions?

Yes, sir. [OPERATOR INSTRUCTIONS] Your first question comes from the line of Ross Seymore with Deutsche Bank. Please proceed.

Yes, hi. This is [Sidney] for Ross. Thank you for taking the call. George, could you explain what is going on with the HCD inventory a little bit more, especially the enterprise side?

Yes. I mean, enterprise inventories are fine -- we haven't given any indication that we had any issues there, and so we're -- we didn't have any Q3. We do not expect to have any issues in enterprise in Q4. We did see some inventory issues in one of our network -- I'm sorry, notebook drive customers, but again, we think that's primarily behind us and we are getting back into a balance of consumption versus demand with that particular customer. The majority of the issue that we faced in Q3 was really related to the desk top customers that we have. As I indicated, we now believe that problem is behind us.

Okay. So in this quarter, will you be shipping [through] consumption (indiscernible)?

We are indicating a modest growth in that business. So I think that would indicate that.

All right. Thank you.

[OPERATOR INSTRUCTIONS]. Your next question comes from the line of Bill Lewis with JP Morgan. Please proceed.

Great. Thank you. I would like to ask about the Intel acquisition you just closed. Particularly, when do you anticipate you will start to see some gross margin improvement, based on the manufacturing changes that you said, and you updated the dilutions for the December quarter. How about the target for when this will become accretive, if you could kind of review or update us on that?

Sure. Well, let me answer the second one first. The timetable, Bill, of the business becoming accretive to us is consistent with what we said, when we -- you know, announced the acquisition which is, it should be accretive to us in Q4 of our next fiscal year. We believe that's the path that we are on. The gross margin should improve slightly every quarter with probably the most first significant improvement coming in Q2 of next year.

And so just as a clarification, you gave a $0.05 dilutive estimate for January, and then another $0.02. Is the $0.02 in addition to the $0.05?

Yes, it is.

Okay. Thanks.

And your next question comes from the line of Michael Masdea of Credit Suisse. Please proceed.

Yes, thank you very much. Just to follow up, Sehat, on the wireless piece, what is the -- what do you think the issue was historically with their business in terms of not getting as much penetration as we would have liked. And what is the customer relationship side like? Is that an important piece that you got with the acquisition or is that something that you have to build up?

Okay. Michael, so if -- as I'd mentioned earlier, the biggest issues is really related to the manufacturing issues. I mentioned about the fact that some of the products -- I mean, the products are built in lower volume, manufacturing lines. The products are built in 8-inch facilities instead of 12-inch facilities. I mentioned earlier about the cost issues. So the fact that it's also in the lower volume manufacturing lines also caused an issue with supply constraints. I mentioned also earlier that there was -- historically the business has always been supply constrained.

So if we look to the history of the business, there were a lot more people that wanted to use the product, but because of the supply constraint issues, remember, the (indiscernibles) have been used for other products, including chip sales and flash. So when there are certain businesses that requires the -- chip set requires small capacity, I mean, the sales of business will obviously have to take a hit in terms of the -- on the supply side. Now that was the past. Okay?

This is improving as a lot of chipsets are moving to different paths, as well as down the road, as we move up these cell phone products into our foundry, this issue of the capacity constraints will even more resolve, as well as not mentioning that we will build this in 12-inch fabs. So the cost structure will be significantly better than running it on this low volume captive paths with a process that is actually optimized for performance, instead of optimizing for cost.

So -- again, so the issues is really in the past. We believe after the next -- in the next one year, we will be able to resolve a lot of these issues and the benefit -- the beauty is the majority of the efforts, the software efforts, the platform efforts, that thousands and thousands of -- many years of investments that each of the customers have placed into this architecture will live.

When we move to our foundry solutions, there's no -- our customers do not have to change anything. All they have to do just -- just test the new chips. And if we change the (indiscernible), at most we will be just changing the boards and then the software will be functioning as is. So that's even more important. This is the reason why we look at the -- when we looked to acquire this business, we look at the future. We don't look at the past. We are comfortable. I'm comfortable that the past is the past and we can solve the future.

All that technical stuff, it makes sense. But what about the sales side? Do you feel like you have the right customer relationships, the right salespeople in place and understanding who determines the design wins and who gets into what phone? Do you feel you have that piece down?

Oh, yes. Yes. That's by default. When we talk to the -- when we talk to the customers, we go to visit the customers actually, before closing the business. I was blown away by the acceptance of the technology of the customers. If the customer ever complained to us, at the time, it was the capacity issues. It was not a single complaint about the technology issue. They all told us, clearly, the XScale technology is years ahead of the competition. If just they have the higher capacity and lower cost solution, things would have been a lot different. If there were no issues with the cost issues or the capacity issue, we would not be talking about this today.

Thank you very much.

And your next question comes from the line of Chris Caso with Friedman, Billings, Ramsey. Please proceed.

I thank you. George, I wonder if you could provide some update on the dilution from XScale that you see for fiscal '08. I think originally you were saying that was going to be 5% when you originally did it. Where is that right now?

Yes, Chris, I think at this point, we are not going to talk about next fiscal year. I think in answer too, I think it was Bill's question, we still are on a path to have the business be profitable, and therefore accretive to us by Q4 of next year. So, you know, that would imply probably for the year we have -- it will still be dilutive, but I think we'll wait until after the Q4, to get a little further clarification on that.

Okay. Well, let me try something else then. With regard to the hard disk market, could you give us perhaps an update or shed some light on some of your plans on potentially growing share in that market? What do you think you guys can do? And where do you think you may be, in terms of engagement with some customers to gain some share in that market?

Sure. Well, again, you know, we have to be very careful, especially in these types of forums about talking about our customer, our customer relationships. I think it's clear that everyone, I think, understands what our market position is and what our objectives are to try and continue. We have been very up front that we have an objective to be a supplier to all the major players in the market. We are moving very aggressively to that.

I think Sehat has made those comments as well. Those things will be -- we are looking forward to updating you folks, hopefully during the next few quarters as to the success that we will enjoy there.

Yes. In my comments earlier, I mentioned that we continue to accelerate our investment into the -- into this business, so we are putting a lot of attempts, technologies to further differentiate our productions against our competition. Our goal is basically to allow our customers -- the customers that are using our technology -- to build products at the lowest costs, at the lowest total ownership costs. And we have been -- we have been able to, over the years, over the last few years, five years, we have been able to prove that any single customer that uses our technology, actually, have better cost products, cost of ownership products, compared to the competition. And what we need to do is -- what we are doing is to further implement the newer technologies and we are already doing that.

We already built some of -- for example, some of our latest solutions that we built. The 90-nanometer designs -- we actually built the technology about two and a half years or so ago. So we are ahead also there, and introducing new features that will allow our customers to further cost-reduce their manufacturing operation costs. And increase performance, among other things. So we are confident -- we are confident that with all of these efforts, we will continue to gain new design wins over the next year.

Okay. Thank you.

And your next question comes from the line from Seogju Lee with Goldman Sachs. Please go ahead.

George, just in terms of the Intel acquisition, can you update us in terms of the numbers of employees that are joining and how we should think about the operating expenses?

Yes, as we indicated, the business had about 1400 people associated with it. Those were primarily the engineering folks. There were also enough, as we worked through the business from June to the closing in November, we worked very diligently on making sure that we took the right engineering folks. But also it became very apparent that we needed some additional support in the sales and the marketing area for this business. So that did end up increasing the costs above the original thought that we had, because of those additional functions that we're taking on. Overall, it was around 1250 people.

Okay. Great. And then in terms of the dilution, the $0.05 -- that is excluding employee stock options. Just to clarify, right?

Yes, I mean, it's just a cost of Intel. Yes.

So that's non-GAAP pro forma OpEx.

Yes.

Now in terms of the options issue. I know you are somewhat limited in terms of what you can say but how should we think about the milestones coming up going forward?

Sure. Again, we can't really say any more than we've publicly said, except that we have a parallel process going on now where our independent -- --our Special Committee is still working on its review and is trying to get that completed now, as quickly as possible. And we have begun the early parts of the restatement process, such that when this whole thing is completed, we will be able to hopefully move very quickly at that point to be able to file.

Okay. Great. Good luck.

Thank you.

And your next question comes from the line of Louis Gerhardy with Morgan Stanley. Please proceed.

Yes, good afternoon. Just on the supply agreement with Intel, what percent of your plans for the next fiscal year would you expect to come from Intel, versus outside foundries? And then maybe Sehat, if you could comment on the timing, maybe the slope or the growth of XScale versus the base-band business for fiscal '08. Thanks.

Louis, on the first one, I think we had indicated way back, even before the deal, that for the next through calendar '07, we would expect essentially 100% of the supply to come directly from Intel. We will be transitioning during '07 with them from the buying of finished goods to a foundry-type model where we will be purchasing wafers from them and taking on the back end processing ourselves.

So for primarily all of '07 we expect the product to come from them. By the time we get to '08, it will be a blend of their production, and product coming from our own foundry model.

And then the second one?

Can you repeat the second part, Louis?

Sure, the second part is if you compare just the slope of the growth in the next year and maybe the timing of the growth in terms of the XScale versus the base band business.

Oh, yes. Okay. I don't have those numbers, but I can give you a general idea is that, obviously certain customers buy just the application processors and there are customers are buying the whole thing, the whole integrated baseband and application processor. And some customer -- and the third type of customer is buying baseband, plus application processor separately on two chips, because they want even more performance, because effectively, by buying two chips, they have two XScale processors on the platform.

If you look at it, for people who have their own baseband, those are people who are using the application processor first. Now in time, we believe as the more stable in terms of using the application processor from XScale, over time, they will be able to transfer -- I mean, move on to the integrated baseband application processor. If you are looking from the customer point of views, from the market point of view, the consumers, actually, will see that -- will look at the different phones really from the features of the phone, and the features of the phone are differentiated by the application processor. There's a reason why that all of these -- the majority of these high-end SmartPhone devices are using XScale today.

The baseband is something that you have to have to talk to the carrier. And some -- this happened to be some of them maybe have software that already work with existing chips on the baseband, and then those are the one that move to the -- for the application processor first. And they are -- their customer like [rimmed it], ramping the holding in a single shot. So we felt eventually, everybody will move to a single chip solution from us. And some of them are already in the process or doing so.

Thank you.

And your next question comes from the line of Arnab Chanda with Lehman Brothers. Please proceed.

Oh, thanks. I have two questions for either Sehat or George, or Weili. First of all, about the Intel business, I know we are talking about this ad infinitum, the product was really good as you said but certainly they are far behind in the market in terms of the market share from OMAP. What do you think the reason for that is? Is the unique professor strategy that you have, do you think that's a differentiator?

And I have a follow-up about the optical storage business. Is there a market entry situation where before you get an HD Blueray, you are not going to get that launched? Or is that something that we see in the back half of calendar '07? Thank you.

Let me take the first one.

Yes.

Arnab, I think it's, again, a misperception. I think Sehat pretty much already went through it. This is not an issue of Intel not being successful. This is a limitation on the amount of product that was available to come into the market. They -- they sold every chip they could make. And it's our belief that this year we will sell every chip that we're going to be able to make under this agreement. So there's no issue about the acceptance of the product by the customers. It purely now, as Sehat has pointed out, is a manufacturing strategy issue that we have to put into play here during '07 to, number one, meet the very nice growth projections we already believe are there for XScale business but to really make it very much bigger in calendar '08.

Yes.

And I want to add a little bit on this OMAP. I think that is unfortunate. There are people that are trying to send the wrong message to the market to take advantage of the fact that Intel did not talk too much about this business. And the truth is the majority of the SmartPhones are using XScale, they are not using OMAP. The OMAPS were shipping to the higher volume, that's true, of the lower end phones, so that's -- that's what happened in the industry. And as we resolve the issues of the cost structures and the capacity constraint issues, we will go into higher volume, mid and SmartPhone markers as well.

And in the long run, there's no reason why we won't go into the lower end segment. It's just a matter of time that we will go into lower end segments as well.

The other question was, will the optical storage penetration be limited by Blueray?

Oh, the answer to that is probably somewhere in between. The market for the DVDs will still be very, very large. Nobody is projecting any disappearance of the DVD probably until the end of this decade. So it will not happen overnight.

Now, at the same time -- so we will be building both, not just the straight DVD only, but we are also building the Blueray and the HD DVD solution. In fact, we are internally busy working on several different versions of the chips that will integrate the Blueray and the HD DVD capabilities on top of the existing device.

So the main difference between the solutions really is a couple of things at the back end, the digital back end, meaning like the ECC -- the error correction -- the format, the formats are slightly different, but everything else on the front end is pretty much the same. The software will be additional on top of, of course. What it means exactly is more work. But, fine, that's life.

Things get more complicated, okay, so we will have to put more resources for those DVD and HD DVD solutions. But we believe that we will be successful not just in the Blueray and HD solutions but in the existing market, we will be -- we will get some design wins as well. Our solutions are based on events, DVD technology, so there are a lot of issues there that people are complaining about at home, about their DVDs, not be able to read some other disk that was written in some PCs. All these issues will be resolved using DSP technology.

So we are confident that the superior solutions will get the design wins, especially for the higher quality end brand -- better brand products.

We might not be able to win the lowest, lowest end, bare bones -- lowest cost products that don't care about quality. So fine. So be it. Those are the markets that we will not address. If we have to miss that opportunity, we will miss that opportunity. We will try to only -- we will address the players that care about performance first. And I think the market is big enough and we don't have to do -- we don't have to drop our shorts to get the lowest cost products. They don't care about the quality. And we -- we are comfortable with that.

Thanks, Sehat and George.

All right. Thank you.

And your next question comes from the line of Jim Liang with Cowen and Company. Please proceed.

Oh, thank you. Can you talk a bit more about your strategy in terms of the integrating the acquired Intel business, specifically in terms of blending potentially two different cultures together? And also, talk about any potential area of product synergy, specifically on the -- when you talk to your SmartPhone customers, any early feedback as far as bundling the XSCale processor with your own Wireless LAN and Bluetooth going forward.

Yes, I will talk about the culture.

Sure. Yes.

Okay. So here's my view about the culture. When Marvell was -- when I started me, Weili and Pantas, we started Marvell, we realized we have to build our own culture of course. But at the same time, when we look around about what are the -- which cultures are the ones that are the most successful in the history, we actually looked at other companies, other big companies that are successful.

So we look at the culture of Intel, the culture of HP and we adopt the culture of HP. We adopted the culture of Intel.

When Intel says only the paranoid will survive, we adopted their culture. The HP way, we look up and we asked around, what is the HP way? We adopted the HP way. So in a sense there, we -- our culture is not much different than the cultures of Intels or HP. You know, maybe as the company gets bigger some of the people and some of the groups are sort of forgetting their own culture, but, this is not an issue to me. I believe it is not an issue.

After all, we come from the same background. It's a matter of sitting up, sitting together and realizing, okay, that we are all the same. We just -- we're -- okay, we're not -- we're not that much different. We just -- maybe we are -- we have different businesses have different issues. We just have to resolve those issues and take advantage, the best of the best, okay. If we have the best -- the best known -- what is that? Design methodology or manufacturing technology or best known practices, we will share it with each other and I think we'll have and be a better company as a result.

And we are moving from that understanding. And I think a lot of people that we talk to are agreeing with this.

I think one of the positive things for the Intel folks joining us is that they are coming back into -- we haven't been around as long as Intel has been around. So we are probably considered to be a little more entrepreneurial at this point, in the life of our Company. So I think from what we have seen so far, they welcome the ability to come in and be part of a little more younger, entrepreneurial environment.

So, we're excited about that.

Yes. And as far as the XScale synergy, I think today for the cellphone space, before the acquisition, Marvell Wi-Fi, and XScale already are on a similar platform in many customer base. As far as the overall product ecosystem, XScale is going to be the platform for all the emerging consumer platforms there for other products that we develop at Marvell. There's a lot of potential.

Right.

And it's not just a chip. It's -- you say, George mentioned 1250 people, the majority of the people are working on the platform and the software. That's huge, huge investment there. And a huge value to our customers. Our customers could buy chips from our competitors, but then they have to -- most of the time they have to develop all the software from scratch. And it's not just the cost of the chip that you have to pay for. It's the additional cost of developing the software, and the additional cost of the tool, the tool chain, (indiscernible) XScale assembled with the most advanced compilers software tools, debugging tolls. This is some of the industry's most advanced tools available in the market.

So the XScale comes with the piece of mind that whatever you invest today will live for the next, I don't know, ten years, five years. Ten years? There's no need to change. Our customers do not have to worry that two years from now, three years from now, if they buy from an application processor from our competitors, they have to change the software, every, every generation, or two generations, they have to change the software. So, you know, historically, people have to use (indiscernible) and then after a few years they have to switch to [army nine]. And then a lot of things have to change, while in XScale, the things that we built are so far ahead of the competition, okay -- and the investment, the commitment that we then put to make sure that, okay, XScale lives forever, okay, it's important our customers.

And the third parties and other third-party software support, that's very, very key. Yes.

And just a follow-up on that point. In terms of embedded Wi-Fi, where the Feroceon professor was a key differentiating technology in terms of getting the significant design wins, do you see some synergy between XScale and Feroceon, or do you see them merging eventually into one?

Yes, as I mentioned, we are actually moving forward. We are combining the resources of XScale and Feroceon team to work together to address the future of the XScale migration. A lot of people -- I guess maybe I need to address what's the difference between, us and the competition in terms of XScale or Feroceon. The XScale team and the Feroceon teams are the only two teams today that are publicly known -- they are publicly known to be developing processors in-house. That utilize the arm architecture, arm instruction sets. Okay? So we are the only two publicly known entities that have this development in-house.

So combining this team -- actually by combining this team together we will now be the only team in the world that have -- that have publicly announced a dedicated, full feature processor team in-house. So this is a big deal, because by -- by having the process architecture in-house, we are able to build -- we are able to incorporate new ideas, new features, new enhancements into our processor product lines, so that we can -- we can get better products, lower products, as well as we can control our destiny that we can give our customers a road map so that whatever they built, whatever they invest in software, or in platforms or in tool chains or in debugging tools, or multimedia softwares, anything they built in the past will continue to live on.

The majority of the costs -- if you look -- if you look for the costs of building products, five to ten years from now, the majority of the costs will actually be the software costs, will be the resources, the software resources to build all the different products. It's the last thing in the world that you want to rely on products that will change every year, every two years. So -- so we believe by bringing the team in house to develop this processor, okay, we will be in the better -- our customers will be in a better position, okay. So it's a long -- I give long answers but it's important to say it. This is not just because we have an ego that we want to build our own processor. It's not. This is something that if you look in the history, okay, this is past anybody would wants to -- to have a long-term product, okay, would have to invest.

Well said. Thank you very much.

All right. Thank you.

[OPERATOR INSTRUCTIONS]. Your next question comes from the line of David Wu with Global Crown Capital. Please proceed.

Yes, good afternoon. Thank you for taking my call. I just wanted a clarification. I have a real quick question. The clarifications in your comments, you said you are going to resend some of the existing designs to presumably lower cost solutions. Do the customers have to take the new designs and go test the software and also work with the operators before they can use your existing designs? The other question I have is really when I look at Marvell'senry into the business, I was wondering what your patent position is in terms of royalty negotiation between yourself and Qualcomm we have seen examples that Qualcomm is fighting in terms of patents and IP negotiations.

Okay. I will address this. The first question first. And then George will take over the -- the patent issue. Okay. So -- what was the question again?

Oh, about the redesign chipped and the qualifications.

I got it. I remember now. Yes, in terms of the -- in terms of some of the redesigns, we are already working on, in progress, the goal is to have the chips to be exactly identical, not just in pinout, but in -- not just in the software but also in the pinout. So the -- the -- their activities going on to the nicest -- we are short in terms of capacity to have devices that use -- devices that use same design, just forcing the libraries of even maybe doing opticals, optical advising or shrinking of the device.

Now in the long run, obviously, there will be -- there will be additional optimization that we will do, but those will be the second phase where we will -- we will further improve certain features or maybe add new features while at the same time, supporting the foundry. So at least for the first space, our customers will not see -- will see no difference in the software. In the future, they will be deciding more features so there will be additional software if they wanted to. If they want to add -- if they want to turn on the capabilities then they have to add software, obviously.

So the second part of the question, David, obviously this is not the forum to go into a detailed discussion, on, you know, patent portfolios and so forth, but I would like to point out that when we made the press release, announcing that we there completed the acquisition, we did mention in there, you know, that we got a substantial number of patents came across, included in the purchase price for this business. So I think, you know, we're -- our patent position as a company continues to be very -- very strong. And I'm sure we will have a discussion with other, people in the market, and leave it at that.

Okay. George, just a clarification. The design is really to increase the capacity, not to reduce costs, correct? Because the pinout is the same.

It accomplishes both, Dave they accomplish both. It's the opportunity for the same design to do somewhat certain percentage of optical strength because some of the technologies in the foundry, actually, -- they see the advanced process technologies. [inaudible -- heavy accent].

Your next question comes from Allan Mishan from CIBC World Markets.

George, can you repeat the gross margin for the October quarter and then I have a question about wireless.

Sure. So basically, we -- we kind of -- you want the actual numbers not guidance?

Right.

So initially we had on our conference call from Q2, you know, put out a -- we can't get into specific numbers but we had indicated product mix was going to be a little -- a little less favorable than it historically was and therefore it was likely that our gross margin in Q3 would be down percentage-wise, would be down from where it was in Q2. Then in October, when we updated for the -- for the revenue, we indicated that, again, the product mix -- we would probably be a little bit more unfavorable from a gross margin standpoint and that we would have some absorption issues based on the lower volume that would cause a further pressure to the gross margins. So without getting specific number, the revised gross margin expectations were down from the original ones on a percentage basis, and then the commentary was we basically came in at that number.

Okay. And then on the Intel business, I think Sehat was saying that most of the improvements on costs will come from the manufacturing side and then I think it was you saying that most of what will be supplied in '07 will actually be from Intel. So how do you get accretive by the end of next year if you are still making everything at Intel and most of the benefit will come from manufacturing?

Well, I think there's two things. One is the -- there's another generation of chips that have already been designed by Intel, which come into the flow. We believe in Q2, which is consistent with when those newer versions of the chips will be actually available for sale. They currently exist and are in the qualification process right now with the -- with the customers. But they come into the sales mix which they have definitely a better gross margin profile than the current chips being shipped. That's number one.

Then number two, as I indicated, we will be moving away from what we call a turnkey model of buying finish product from Intel to a foundry model, putting them into a typical foundry model where we will do wafers and that historically is also more beneficial from a cost standpoint. So those two things along with the increase of volumes, and the absorption benefits of that are what's going to drive it towards accretion in the second half of next year.

Okay. Thank you.

And your next question comes from the line of Adam Benjamin with Jefferies & Company. Please proceed.

Yes, thanks. Just two quick questions. First on embedded. Your commentary was that embedded would be down in Q4. I think you were referring specifically to the gaming, but can you talk about the wireless LAN as a total, including embedded and whether you expect growth sequentially in the fourth quarter and then what are the drivers there back and forth? Because you would expect given the gaming ramp in Q4, you would see some growth there. Just trying to reconcile that, thanks.

We are an integral part of the gaming platforms and that's very exciting for us. I think it's pretty publicly known that the product that will be built on those platforms this year, you know is limited. And so, you know, we're -- we have to face that as the reality that we are seeing. So we are not really anticipating, you know, in the fourth quarter as opposed to the third quarter on those platforms. And then some of the older platforms that have been in the market actually are seeing a downward revision in their production rates for this time of the year. So that, is the issue or explanation on gaming.

Overall, yes, wireless LANs should do fine. Our other areas of embedded are quite, you know -- are doing quite well. So, yes, we should be quite fine overall. We wanted to point out that that is the situation because that's been such a highly monitored platform.

Right. Okay. And just a follow-up on Intel. I know this has been asked a couple of different ways. In Q4, you are looking for $90 to $100 million, the gross margin you have talked about is roughly in the 20% range and the OpEx is about $50 million for that business. So roughly, you are going to lose maybe $30 million in Intel in the quarter. How do you get from $30 million lost to profit by Q4 of next year. Can you break it down as to the combination of the revenue contribution or is it a reduction in OpEx or margin? Just quantitatively, on a percentage basis which is contributing the most there?

Yes, you know, I think we already said Adam, we are not going to touch next year at this point. We certainly, after Q4 is over, we will address, fiscal '08. I think, though, we have already said it previously, all of the items which you raised will be positively impacted as we go over the next year, including an increase in the revenue ramp, a decrease in the product costs and an optimization of the overall operating expense structure of now the new combined Marvell/XScale. So these are all things that we previously said that we will -- we will, and eventually already have and we will be continuing to work on going forward.

Yes. And as our customers migrating from the older products to the newer products, some of the older products which is worse on the margin side, will have a better margin on the newer products.

Yes. Right. But as you see that and plan for, that you can't give any kind of contribution level of which one is going to contribute more to get to that profitability?

We never do that anyways. We're not prepared to do that today and that is not something where we get to that level we will -- once we are in a position to talk about our full financials, we'll give overall gross margin expectations for the company. We'll give actual operating expense trends. We're not in a position to do that right now.

Thanks.

Ladies and gentlemen, that's all the time we have for questions today. At this time, I will now turn the call over to Sehat Sutardja for closing remarks.

Thank you Letitia. This completes our Q3 fiscal year 2007 conference call. I would like to thank everyone for joining us and look forward to the next quarter.

Thank you.

Thank you. That will conclude today's conference. That concludes the presentation. You may all disconnect and have a good day.