邁威爾科技 (MRVL) 2004 Q2 法說會逐字稿

Ladies and gentlemen, thank you for standing by.

Welcome to the Marvell Technology Group Ltd. second quarter 2004 earnings call.

During the presentation, all participants will be in a listen-only mode.

After the presentation, you will be invited to participate in the question and answer session.

Conference call is being recorded on Thursday, August 21st.

I will now turn the conference call over to Co-Chairman and Chief Executive Officer of Marvell Semi conductor, Dr. Sehat Sutardja.

Thank you, Matthew.

Welcome, everyone to our second quarter fiscal year 2004 conference call.

Weili Dai, Executive Vice President of the communications business, and George Hervey, Vice President of Finance and Chief Financial Officer, are joining me on this call.

Today we are reporting our quarterly earnings approximately one month after most of our peers have reported.

From those reports, most of us in the industry have heard the conscious optimism regarding the state of the economy.

In general, those companies they are addressing new market developments are more optimistic than those that are just addressing the existing markets.

This should not be surprising to many of us that have followed the market as well as those who have followed the Marvell story for the last several years.

As you may have guessed, we are not going to buck the trend of the caution optimism on the general economy conditions.

However, as we continue to develop, many new semi conductor and system solutions for our customers, we are optimistic that we will continue to reap the benefit of getting a larger piece of the new business opportunities from our customers.

Despite the challenges that we face every day during the prolonged economic downturn, we have actually accelerated our R& D investments into several new areas as well as in all of our existing market segments.

As the industry will eventually go back to a more normal growth rate, we will benefit more from these investments.

Before I go over some of the highlights of our current businesses, I would like George to give you the safe harbor statement and the review of our Q2 financial.

Thank you, Sehat.

Good afternoon, ladies and gentlemen, I'd like to remind all participants that the following dialogue will contain predictions, estimates and other forward-looking statements covering subjects such as: data storage and communications market trends, competition, customers, suppliers, products and demand, 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 other words with similar import.

These statements include those relating to the pace of our business as we have completed our first and second quarters and the impact of the continued adoption of our solutions on our revenue growth.

The following factors among others could cause actual 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, increase 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 10k, recent quarterly reports on form 10q, recent current reports on forms 8k 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 conditions.

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

Moving onto the Q2 financials.

Today, we reported net income -- I'm sorry.

We report both net income, loss and basic and diluted net income loss per share in accordance with GAAP and additionally on a non GAAP basis referred to as pro-forma.

We believe the non-GAAP information is useful because it can enhance the understanding of the company's ongoing economic performance and Marvell therefore uses pro forma reporting internally to evaluate and manage the company's operation.

Marvell has chosen to provide this information to investors to enable them to comparisons of operating results in a manner similar to how the company analyzes its operating results.

Today we reported that net revenue for the second quarter of fiscal 2004 was a record $192.9 million, an increase of 61% over the $119.7 million reported for the comparable quarter in fiscal 2003.

And a sequential increase of 15% from the first quarter of fiscal 2004.

Pro forma net income which excludes the effect of acquisition related expenses, amortization of stock-based compensation, and charges related to facilities consolidation was $30 million.

Or 22 cents per share diluted for the second quarter of fiscal 2004.

Compared with pro forma net income of $14.2 million or 11 cents per share diluted for the second quarter of fiscal 2003.

Shares used in computing pro forma earnings per share diluted for the second quarter of fiscal 2004 increased to $136.8 million as compared to $129.6 million shares for both the first quarter of fiscal 2004 and the second quarter of fiscal 2003.

This quarter, we'd also continued to increase our net income under Generally Accepted Accounting Principals--GAAP.

Our second quarter fiscal 2004 GAAP net income was $9.4 million or 7 cents per share diluted compared with net loss under GAAP of $9.3 million or 8 cents per share diluted for the second quarter of fiscal 2003.

This represents our second consecutive quarter of GAAP profitability since our acquisition of Galileo technology in the fourth quarter of fiscal 2001.

We have provided on our web site in the investors section at www.marvell.com a reconciliation of GAAP net income loss to pro forma net income for the quarter, reported today plus the prior eight quarters.

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

Thank you, George.

In this conference call, I would like to give you a different view of our company.

Specifically, how we built our businesses focusing on two very important topics.

Our business direction and philosophy.

In the past, I have mainly talked about our technology superiority and the countless list of products.

After a while, people have started to think about us the technology machine that just happened to be early in the game and that's about it.

In business, after all, our customers don't just buy the best products that they can find.

If this was the case, we would have already captured the majority market share in every market that we entered.

Over the years, we have learned that many of our potentially large customers actually have other requirements for doing business, besides just looking for the best product solution.

Many of them are looking for company that they can work with for a long time.

Switching a key critical component supplier can be very costly and disruptive to the normal operation of their business.

Many of them are also looking for suppliers that they can trust to work with in good times and bad times.

As a result, they are always looking for a stable supplier that will continue to be in business for a long time to come.

Why am I talking about this?

Well, this all has a lot to do with building a company that will provide long-term shareholder growth and value.

The founding members of Marvell, myself, Weili, and Pantas have from the very beginning had the philosophy of building Marvell to be a long-term growth business.

We always question ourselves how can we achieve this long-term goal in everything that we do at Marvell.

Our long-term goals have always had higher priority than short-term goals even if it means letting go of short-term business opportunities or initially taking a hit on supporting a small opportunity that we believe could be very successful in the future.

A key guiding factor in our decisions in how we can -- is how we can help our customers to be more successful.

So that we both can mutually benefit from our businesses.

We also understand that as we are dealing with large businesses, management interaction between supplier and customer can become time consuming.

This interaction provides us the direction for our future growth as we build more complete chip and system solutions for our customers so that they can limit their involvement with fewer suppliers.

Our motto to our customers is always do whatever it takes to make them more successful.

Once we commit to a customer, we always stick with our commitment.

Now the philosophy-- we never go into a business just to be a need to supplier unless we believe we can completely displace the current leader in the market, we will not invest our precious resources to start any new development.

Even after all these criteria are met, we will only go after business that as a whole offers large potential revenue.

Once we decide to go into a new market, we will wholeheartedly invest ahead of the revenue stream to make sure that the market takes off and as a result, we hope to benefit disproportionately from the success.

Finally, we are in the business not for the short-term gains but more importantly for long-term success.

With long-term success, we hope to achieve the long-term financial gain.

When we combined our business philosophy with the best talent and technology foundation, we expect success.

An example of this is the Gigabit ethernet.

When we entered this business three years ago, nobody would have thought that the Gigabit would replace fast ethernet in the business environment.

By driving the technology for the Gigabit while at the same time working closely with the leaders in the enterprise market, we have successfully moved the markets more towards Gigabit ethernet.

Today, the desktop client transition to Gigabit is running full steam ahead while the transition on the enterprise has already taken off.

Our commitment in this area is obvious.

While we originally concentrated our resources at the enterprise level, we now are also addressing the entry levels, low cost Gigabit switch markets with our latest products.

We believe by the end of next year, the majority of the enterprise solutions from the very low end to the high end markets will be Gigabit [safe].

Another example of our successful strategy is our storage business.

This business is a classical business model for which all other future business that we target are compared against.

We believe it is now just a matter of time when we will become the leader in volume shipments of storage and storage-related components for this drive.

Our commitment to this business through industry-leading technology and cost effective solutions has proven to become the driving factor of revenue growth in a business that otherwise has continued to see significant cost pressures.

We believe the eventual large market opportunity beyond the [PC] market, i.e., the consumer market is the single biggest reason to continue to invest heavily in this technology for our storage business.

It is not hard to imagine that in two years' time, every home will have at least one or two pvr's or personal video recorders incorporating the hard-disc drive base storage device.

It is also not hard to imagine that all high resolution digital cameras will come with a microstorage device in the near future.

But closer in time, we will see a huge [transition] to Prestera ATA, [inaudible] SCSI in the desktop and enterprise market.

Two years ago we pioneered the development of the industry first Sierra ATS solutions resulting in many design wins.

We have just recently introduced the industry first next generation serial APA2 solution.

When combined with our SoC system-on-chip solution, [inaudible] technology will be another driving factor for gaining a bigger market share in this business.

By replicating the recipe for success that we have today into new areas, including on a lot of things, wireless LAN, we hope to grow the company into one of the largest semi conductor suppliers in the world.

Recently, I found out that we already are one of the top ten largest [inaudible] semi conductor suppliers in the world.

My goal is that within three years' time, to move out into the top five suppliers.

Unlike many other top suppliers in the market, our business is more diversified.

And as such, we are not too dependent on the up and down of each businesses or the market commission for our success.

Furthermore, all of our businesses that we address are individually large in market price with each one of those businesses potentially reaching a top ten supplier status.

And hopefully after all the hard work, we, the company, our customers, and our investors will benefit financially for the long term.

Now, I would like to turn the call back to George for additional comments regarding our Q2 financial and forward guidance.

Thank you, Sehat.

I'd first like to make some additional comments on our Q2 results.

The Q2 revenue of $192.9 again was a new quarterly record for the company.

And a 61% increase from Q2 last year.

Additionally, the 15% increase in revenue from Q1 to Q2 compares very favorably to our guidance of a 10% to 12% sequential quarterly increase in revenue.

The Q2 sequential increase of 15% represents the seventh consecutive quarter that our sequential revenue growth has been greater than 10%.

During Q2, both our storage and communications businesses performed well resulting in the increasing revenue for both businesses.

Entering the second quarter and reflected in our Q2 revenue guidance was the anticipation that while we remain cautious regarding the overall economy, and a potential recovery in IT spending we exited Q1 with significant momentum generated by our strong market position for our product portfolio and increasing number of designing wins for those products.

That resulted in our visibility for Q2 improving when compared to our visibility entering Q1.

As we progress through the quarter, Q2, those factors remain consistent with our expectations.

Additionally, during Q2, we saw increasing demand in the mobile storage market and the increased adoption of Gigabit ethernet by the pc client driven by the launch of SpringDale platforms in May 2003.

All of the other fiscal 2004 revenue growth drivers continue to increase consistent with our expectations.

These include desktop storage, wireless 802.11, Gigabit ethernet network infrastructure and Serial ATA.

For the quarter, storage products contributed mid-50% of total revenue with communications products representing the balance.

Q2 gross margin of 53.9% was essentially in line with the midpoint of our guidance of 54% plus or minus 25 basis points.

We continue to experience outstanding manufacturing efficiency and the benefits from our yield improvement programs continue to produce very cost effective products.

This has allowed to us achieve our individual targeted product margins.

Q2 product mix which is the largest determiner of gross margin percentage was consistent with our expectations.

As we've mentioned over the past quarters, we're committed to continue our level of investment in our new product development and business structure as we position Marvell to grow to the billion dollar revenue level.

While accomplishing this objective, we are also committed to have our pro forma operating expense decline as a percentage of revenue.

In Q2, our pro forma operating expense as a percentage of revenue declined to 37%, which represents a 200 basis point decline from Q1 and 100 basis points favorable to our Q2 guidance of a 100 basis point decline.

For the eighth consecutive quarter, we increased our pro forma operating income percentage.

Our continued double digit sequential revenue growth balanced product mix generating consistent gross margin percentage contribution and decreasing pro forma operating expenses as a percentage of revenue resulted in 120 basis point increase in Q2 pro forma operating income percentage, from 15.7 to 16.9%.

The result of this continuous improvement has brought us very close to our long-term model of 18%.

Shares used in computing pro forma net income per share for the second quarter increased to $136.8 million, compared with $129.6 million for the first quarter of fiscal 2004.

The increase in the shares was largely driven by the use of the higher average stock price for Marvell shares and the treasury stock method calculation as well as shares reflected-- related, I'm sorry, to the RADLAN acquisition which closed on June 28.

Moving to the balance sheet, the strength of our balance sheet continued to improve in Q2.

We increased our cash and short-term investments by approximately $45 million to a total of $344 million.

Year-to-date, we have increased our cash and short-term investments by $79 million.

DSO's for Q2 were 48 days consistent with Q1.

Our DSO's continue to remain in the range of our guidance of high 40's to low 50 days and DSO's are likely to remain at the high end of our range during this period of rapid revenue growth.

During Q2, we increased inventory by approximately $14 million dollars to support the current and projected growth of our business.

Our days of inventory increased from 55 days to 61 days but our days of inventory continue to remain below our target of 65 to 70 days.

We'll will continue to monitor our production levels going forward with the goal of putting us in the most favorable position to respond to increases in demand for our product.

Now, turning to the future.

We just completed the first half of fiscal year 2004.

We'd like to take this opportunity to update our guidance regarding the financial outlook for the company for all of fiscal 2004.

During our Q2 call we mentioned that our visibility into how our major revenue opportunities would contribute to our fiscal 2004 -- I'm sorry, during our Q1 call, to our fiscal 2004 revenue improved.

These opportunities include Gigabit ethernet, for pc client, desk-top storage, Gigabit ethernet for the network infrastructure, 802.11 wireless and Serial ATA.

Additionally based on current trends that we have seen develop during Q2 we are adding mobile storage to our list of major revenue opportunities for fiscal 2004.

Our revenue for the first half of fiscal 2004 totaled $361 million coupled with consistent visibility for our designs that are currently in production provides us the opportunity to achieve more revenue growth than we anticipated after Q1.

Based on all these factors, we now expect our revenue for fiscal 2004 to range between $805 million to $825 million.

Which at the midpoint is approximately a 62% growth of fiscal '03.

For the balance of fiscal '04, we believe that our gross margin percentage will remain above our long-term model of 52% but below the Q2 level of 53.9% and R & D and SG & A, as percentages of revenue should continue to decline.

Now, moving more specifically to Q3, '04.

While we remain cautious regarding the overall economy and a potential IT recovery, as I mentioned in my discussion of Q2, we exited Q2 with significant momentum generated by our strong market position, broad product portfolio, and increasing number of design wins for those products.

This has resulted in our visibility for Q3 remaining consistent when compared to our visibility entering Q2.

We have targeted company revenue for Q3 to increase 10% from Q2.

As Marvell continues to grow our quarterly revenue at a double digit sequential rate, we believe that a significant portion of that growth will be derived from the pc market.

This will impact the product mix of shipments in Q3.

While we believe that normal ASP declines will be offset by continued reduction in our manufacturing costs, the anticipated products mix of our shipments in Q3 should result in Q3 gross margin of 53% plus or minus 25 basis points.

With our focus set on providing the technologies, products and building a company structure to support long-term growth, R & D and SG & A, in absolute dollars will increase in Q3 but decline approximately 130 basis points as a percentage of revenue from Q2.

Interest and other income should remain consistent with the Q2 level.

Our shares used in computing pro forma net income for Q3 should increase to $139 million from $136.8 million in Q2.

Now I'd like to turn the call back over to Sehat.

Thank you, George, that completes our commentary would you please poll for questions?

Certainly.

If you have a question at this time, please press star then the number one on your telephone key pad.

Again that's star then the number one if you have a question.

We will pause for just a moment to compile the q & a roster.

And your first question is from Michael Masdea with Credit Suisse First Boston.

Congratulations on another great quarter, guys.

First question relates to what you said about, somewhat of the shift in mobile that we're seeing out there.

Can you talk about how your share is on mobile versus some of the other areas?

Also talk about the--how your seller opportunities for drive is and also the profitability relative to the other segments of storage?

You want to cover it?

Yes.

We have been making significant progress, I think as we've discussed previously in that market.

Traditionally, our two largest customers there had been Toshiba and Hitachi and we're very pleased to add Fujitsu as a major mobile customer for us here in fiscal 2004.

So our feeling now is we are probably approaching 75% market share of the mobile market.

We pretty much down the path of converting those platforms to system on the chip, which helps us obviously from an ASP standpoint and we've discussed before, mobile ASPs tend to pretty much be in the middle range of where our overall ASP's, you know, end up.

So no material -- is there any material difference in profitability versus the other segments?

No.

I mean, you know, everything is -- when you have SOC's or they're competitive.

I would say there is a small difference between the segments but nothing material.

And Sehat, you talked a lot about kind of the long term of the organization.

Can you give us an idea and maybe George, too, what percentage of the R & D today is going towards opportunities which really aren't generating a lot of revenue today?

And also, how long does it typically go or take to go from a lot of those R & D dollars to revenue dollars?

Sure.

I will say I don't have the number exact number percentage, but my gut feeling is somewhere between about half of our R & D's are targeted for future opportunities and some of the other half obviously some closer in range to be like a year down the road.

But some smaller percentage would be like two years down the road.

We even have R & Ds that we targeted at three, four, five years down the road.

Usually on a smaller percentage basis.

But we do have significant R & D investments and the reason for that is because we believe that we need to build complete system solutions for our customers.

So they can shop from us instead of going somewhere else.

Thank you.

Last question just as you guys continue to grow at the rate you've been growing, what changes do you need to make for the organization?

Are there any structural changes or any major changes you need to undertake that we should be looking for just so you can handle more and more revenue?

George, you want to talk about that?

I think we've been on a path now for well over a year.

You know, because again, we've been growing pretty healthy for almost two years now.

We are, you know, trying to bring more focus to the various market segments that we have to address within the businesses, so Weili has been constantly adding capabilities within her businesses and so has Sehat.

So overall, we're expanding trying to bring more seasoned managers into the company as we grow and just build a more solid organization -- we're very good, but we do need to add capability as we continue to grow.

And we want to be focused moving forward.

So I think the head count if we exclude the people from RADLAN, I believe we added 120 people to the overall head count during the last quarter and picked up an additional 160 people with the completion of the acquistion of RADLAN and we're almost, I think, approaching 1600 employees now worldwide so we clearly are continuing to, you know, staff for future growth.

Thanks.

As a reminder, if you would like to ask a question, press star then number one on your telephone key pad.

Please limit your questions to one per caller.

And your next question is from Jim Liang with Pacific Equities.

Thank you.

Can you give us an update on the [integration] of RADLAN and also the product development effort over there?

It [inaudible] the industry of RADLAN is [inaudible] with a portion of the technology, we already accelerate the [inaudible] activity of our [customers] will be able to offer silicon as well as software solutions [inaudible] so absolutely it sets the time to march.

Great.

Just a follow-up.

George, did you say that the pc storage customers are now 100% in terms of SoC ramp?

No, I said the mobile pc platforms have, you know, by and large converted to SoC.

There are still a few hold outs but the vast majority have converted on the mobile storage platforms.

So [inaudible] is the only one that's remaining?

That's not a mobile, that's a desktop.

On desktop side.

Thank you.

Great quarter.

Okay, thank you.

Your next question is from Ambrish Srivastava with GKN.

Hi, guys.

Thanks.

Great quarter.

A few questions.

The first one is George, what drove the upside if you look at the different segments of the business, was it one particular area or was it WD ramping better than expected?

The next two questions, more for Sehat, would you be willing to venture a guess as to how many ports of Gigi infrastructure you have shipped by now?

And you talked about HDTs appearing in digital cameras.

Knowing from somebody who bought a 128 meg flash card paid $90 in the city to get a lot more disk drive space for that.

Any idea as to what timing we are looking at frame as when these consumer drives start appearing in such applications, thanks?

We don't want to get too, too specific Ambrish on our revenue breakout, but as I mentioned in the opening comments, clearly mobile storage is a very, you know, Intel plus positively about it and many of our customers have talked positively about the trends that they see in their mobile storage business.

So you should equate that to it being very positive for us.

And then the other area which is, you know, clearly, it was a lot of confusion and a lot of anticipation as to what would happen with the Springdale launches back in may.

I think as it turned out to be very positive event for Marvell, which started to contribute pretty significantly in the Q2 time period.

Okay the Gigabit five, as I state in our fall, the Gigabit is pretty much moving full steam ahead in the desktop.

I will believe by sometime middle to the end of next year, pretty much all the desktops should be Gigabit based especially for the Enterprise market segment.

For the business environment, the cost difference would be so small that nobody's going to care about saving a few dollars there.

On the Enterprise side, infrastructure, you see for example a recent launch from Cisco talking about Gigabit even on the entry level.

So that's a typical time frame you would see not only from [inaudible] but also from other customers of ours that in the next several quarters [inaudible] products all the way to the low end segment meaning the entry level for the Enterprise.

So the numbers, it's pretty hard to track how many percent because changing rapidly.

The good thing is the numbers the percentage would be increased to 100%.

We believe those numbers are going to grow.

Very significant as Sehat mentioned.

Specifically they now have Gigabit across all of their platforms, that was very important.

Okay, and the last part is the consumers' portable for example, for digital camera.

This is exactly what we've been saying over the last several years, a year or so that storage would be going to the cameras for portable applications simply because of the cost advantage compared to flash memory.

And it's unavoidable.

The rotating media is always going to be cheaper and more cost effective.

Now, in the past this application did not really exist because the resolutions of the pc digital cameras were not good enough.

As the industry moving--going to 4 megapixels to 5 our report just recently for quarter I saw a 5 megapixel camera and then might have by in a year or two is being 8 megapixel coming out.

Definitely, you would need several gigabytes for your high end cameras.

So I -- our theory is you will see a lot more of those [inaudible] device coming out sometime the middle of next year.

Thanks.

Your next question is from Cody Acree with Legg Mason.

Thanks.

Again, let me just echo my congratulations.

Can we go back to Gigabit ethernet for just a second.

I know you said we are starting to get it introduced across a lot of the infrastructure platforms, but how big of a piece of it is it in your revenue stream today?

What kind of momentum can we expect going forward and I guess maybe just a follow-up, of the increase in your guidance, what are the components of that?

Can you break it down at least in generalities?

What's the biggest drivers?

That was a lot of stuff in there.

I think the answer to the first part is that, you know, the Gigabit ethernet infrastructure has, you know, been a little slower to take off than everyone had originally expected.

Q2 was probably the fist time we've seen some meaningful increases there.

From where it had previously been.

But we were very pleased.

Again, I think that goes with what our largest player in the industry has set.

So that kind of signals that from that point moving forward we should continue to see increasing growth, you know, being in Q3, Q4 and beyond should be trending upwards on the Gigabit ethernet infrastructure.

So I think that's consistent with what we've been saying, you know, recently.

As far as we don't want to break down specifically each component, and that's why we group them as revenue drivers, but, you know, all of them are doing better.

And all of them are contributing to the reason why we raised our year target now, you know, again, by another $25, actually $35 million on the high end.

So if there's not one that's dominating it, they are all going to contribute.

Some are a little bigger than others but they are all contributing to the revision in the guidance.

Lastly, can you maybe talk a little bit about your white-box opportunities with the Yukon products and what success you've seen there and is it yet contributing to revenue?

Sure.

I mean, one good example is the largest player the company [inaudible] If you look at the platform, from Aduse across the board 100% using the Marvell Ucom family chips.

So this is extremely successful, and I think moving forward, we'll see more volume and [ramping] to more or their platform.

Great.

Thanks guys.

Congratulations.

Thank you.

Your next question is from Arnab Chanda with Lehman Brothers.

I don't know if I can say marvelous quarter but let's see.

Anyway.

That's okay, can you say that.

Okay, great.

First, can you talk a little bit about a couple of things.

First of all, how much is wireless LAN contributing to you guys and how do you think about that business?

And secondly, when do you think Prestera will start to contribute.

Thank you.

Again, wireless LAN is one of the growth drivers for the year.

We mentioned, I believe, last quarter that Q1 was the first quarter where we had, you know, where we would consider to be production revenue on wireless and, you know, you should expect that, yes it did grow in the Q2 time period and again it's going to grow each quarter, Q3, Q4 so we are excited about that opportunity to, you know, gain more market share there, which we absolutely expect to do as our solutions are more widely deployed.

On the Prestera, Prestera contributions have been very, very minimal, at this point.

You know, the number of customers that are very close to introducing and we expect, you know, that to start contributing again as one of the drivers of the back half of the year.

And we expect that to have its reasonable, you know, percentage of contributions.

And it should accelerate over time because it is a very large market, end market opportunity for us.

Especially the software drivers.

Right.

It's made a lot of progress, you know, technically with customers over the last three months.

And actually, there are a number of customers with the launch product.

Yeah, Nortel is definitely there now.

And this quarter you should [inaudible] And on the one front, Marvell, you know where to not put one that came out with the wireless solution but I think at this point, we absolutely, we are beating terms from a technologies standpoint of view.

Especially we overcome a lot of hurdles, expect consumer markets.

I think how I mentioned last quarter, that's actually moving forward is now a main target.

That's where the volume is.

I'd say stay tuned.

There will be a lot of updates from us hopefully soon.

Thank you.

And then one question about the customers--top customers in the quarter, could you tell us who they were and what percentage they represented?

Thank you?

We certainly will tell you who they are we don't normally say precisely what the percentages were.

Again, Intel remains our largest customer followed by Samsung as the second largest and reflecting again, I think the commentary of the mobile storage Toshiba has now entered a 10 percent or greater list of customers.

Thank you very much.

Your next question is from Jeremy Bunting with TWP.

Thanks very much.

Could you just comment on the relative traction of the Yukon transceiver versus the product which you're selling through Intel?

Are we near a point where we might get some sort of transitions and what sort of time frame do you expect that to be, please?

Thank you.

I have to say both are great technologies.

Obviously, you know, both are Marvell [inaudible].

In the Yukon case we also integrate our own math, but in terms of the transitions I think both products do not get whether or not the Intel promises or the Marvell own penetration overall is a net result of over 70% market share.

That's extremely important.

Okay.

Can you comment on how much of that is contributed through the distribution deal you have with 3 COM?

Well, Yukon initially, you know, with the penetration of the light boxes going 100% through 3 COM.

Okay.

All right.

Thanks.

Your next question is from David Wu with Webb Bush securities.

Have a question on the wireless plan business.

As you move into introducing stand alone wireless plan high-speed products the second half of the year, what plans or what should we be looking for in terms of bundled or integrated products with your other-- your other product lines?

The wireless plans are devised, are pretty much already fully integrated.

It comes with a lot of features so there's nothing much that you need to bundle to build a complete product.

In fact, we are probably the highest integrated products in the market now.

If there are people that want to build even higher [inaudible] like the Gateway routers with wireless LAN, sure, they all, we can bundle let's say a five port [inaudible] switch or an eight port [inaudible] switch or even higher ports that even make sense.

But those will all be a smaller percentage of the customer base.

We get a lot of these not just because of the performance of out wireless plan is superior but is also because of it's very, it's the most integrated device on the market today.

And the example for the Gateway you can buy from the store these days and the [inaudible] and they give us--Marvell a total solution that includes pilots and our software too.

Can you talk a little about I was thinking about actually a one-chip low power version because your Southern California competitor has been talking about imminent introduction of something in the b and one-chip b and they talked about a one-chip g.

If you are going to go into the consumer hand helds, a very low powered whether it's one-chip or two-chip solution seems to be highly desirable.

Are we anywhere close to that in terms of your road map?

Let me answer this question.

Number one, I don't want to make any comment about sales of one chip solution.

The one thing I can tell you is our customers have seen the performance, the noise performance of the device is superior and the power consumption is superior.

This is what you care about when you talking about cell phones.

Types of markets, portable pda's or anything that must operate with battery, battery operation.

Whether it's single chip or two chips or not, to date, regardless how many chips people have we have fewer chips today.

Our chips already integrate power amplifiers, for example, which is conceded to be not a chip for our competitors or anybody else a solution out there.

So when people talk about one chip, they probably forget to tell they also need power amplifiers outside.

In other words, your current generation of b's for example or g's, for sampling or shipping, would be

With one fewer chip basically compared to other people today.

I understand.

But are they suitable for cell phones or pda's?

Yes, because of the power consumption significantly lower powered than the devices that really a lot of the devices out there are optimized for laptop pc or laptop markets where you have large batteries on your laptop versus for pda's where you only have a single lithium battery that only have less than one amp of capacity versus the laptop having 3 cells to 6 cells.

We have five times the capacity.

So the optimizations are different, the portable market device to optimized to be operating on a stand alone basis meaning it has its own processor so therefore the device is more intelligent versus the device they optimized for laptops where they using most horsepower of Pentium to do most of the work.

So our solution's definitely are more suitable for the consumer markets.

Thank you.

Your next question is from Dushyant Desai with CE Unterberg Tobin.

Thank you.

The first question is on the infrastructure for the switching.

If you have any comments about the Asic versus the SSB transition.

The current products are dominated on the switch side with Asic's.

How do you see the SSB solutions like yours being accepted and then ramped up?

Okay.

I will say that for people that always have different opinions on what makes them and there will be those there will think what they only think.

And because of whether they believe better performance or because of the software, they have invested over the last, you know, ten years, and there will always be customers at the same time on the other hand there will always be customers that want to see the lower cost nature of standard products that, you know, we develop.

Either way, we support -- we at Marvell will support both models.

It's really up to our customers to decide whether they want to use Asics or a standard, the beauty is that either way we will get benefits from the business anyway because the majority of the semi-conductor components are actually on the physically aside.

Because forever, let's say we have just recently introduces a 24 port single chip standard products switch so that's a single device for each one of those devices you will need six of the quadport Gigabit ethernet devices.

So either way is fine with us.

And then I wanted to compare the Gigabit ethernet and the serial ATA situation..

I know they are different market segments but, could you give us an idea about the Gigabit ethernet penetration level and where do we stand with the serial ATA penetration level and how do you see that going in terms of the ramp and at some point in time I suppose the serial ATA takes over the majority of the interface?

I don't think I can compare the Gigabit versus serial ATA.

The Gigabit is replacing fast-ethernet so clearly people want to have higher bandwidth using same wirings namely the category 5 cables.

And as long as the cost is not much more, I mean, just a little bit more than the [fast]-ethernet, people will switch to the Gigabit handed down basically to any question with Gigabit especially with the cost seeing now is very, very subtractive.

On the replace of the parallel cable inside the pc, so Gigabit outside the connection is inside the box.

The benefit for serial ATA is huge.

You are basically replacing complex bundles of wires in the ribbon cables with just two pairs of twisted pair wires.

So huge cost savings but more importantly, the performance is increased by 50% even in the fist generation of the serial ATA 1.

That increase in performance actually translates into better benchmark performance on the cpu benchmark.

As we move to the serial ATA 2 next year, we're already into the serial ATA 2 today but we are targeting mainly for the Enterprise first, but we will move quickly to the desktop as well, the performance will double.

The question is will that increase the cost?

Absolutely not.

The cost actually may even go down with a serial ATA.

Because as all the silicants they will build will have fewer pins, I mean lower cost packaging. .

So smaller by size.

So the transition to serial ATA will happen because of that.

And where do you estimate the penetration level for serial ATA?

Is it less than 10% of the drive shipping and then where does it go to?

Probably I know less than 10%, but we'll see probably approaching 50% by sometime next year.

One of the things we've seen since the Springdale, you know, launch in may is input back from our customers to expect higher volume demand in the second half.

For serial ATA.

As more and more chips set, mother boards which have native serial ATA part of it, you know, will be more prevalent in the market.

Got it.

Just the last one, George, on the mobile disk drive, you know, you saw a lot of trends.

How much of that is related to Centrino ramp and Centrino launch versus you know non-Centrino type of either SoC conversion or other mobile device?

I think it's across the board.

You can look at the Centrino obviously is a good driver for increase in mobile acceptance in the market.

More importantly, the lower costs of CST's and other components in the laptop, essentially creating a much better pricing for the laptops.

We're already seeing laptops selling less than $1,000.

So that's really the cost, the lower cost laptop driving the increase in acceptance laptop for the home.

It's mostly demand.

There's really no -- a lot of the conversion had already taken place, so, you know, most of the growth areas were pure demand.

Thanks a lot.

Congratulations on the good quarter.

Thank you.

Your next question is from Karl Motey.

One area you didn't talk about is your Enterprise storage business, on that business[inaudible]

I didn't hear the question, Carl.

Could you repeat it, please?

I'm sorry.

Your enterprise storage business?

Can you give us an update on that and whether or not it grew sequentially?

Okay.

We don't -- one of the things -- one of the challenges of having a company that, you know, getting bigger and more businesses that we address, there will always be times where we don't talk about certain businesses.

But since you're asking about Enterprise, they are fine.

Enterprise is doing quite nice.

The, you know, again, I think the two of our customers who have reported have indicated that, you know, their business is strong and their customers have been reporting, you know, their businesses have been strong.

So we are farther down the food chain but it definitely feeds down to us.

So we're quite pleased with that business, and, you know it's a growing business for us.

Great.

And one follow-up question on the mobile drive business.

I realize you are selling into the 40 gig mobile platforms that are currently ramping.

How long do you expect the 40 gig products cycle that's currently ramping to last?

The 40 gigs for the mobiles actually, the technology is that use the equivalent to the 80 gig for the desktop.

This is the technology that's been quite challenging for the industry to get into.

Product similar to the transitions from .18 micron process geometry to .13 micron technology.

So based on our experience, you now, dealing with this business in the last several years, I would say the 80 gig desktop and 40 gig mobile generation will last a long time.

Through next year and the following year.

Yeah, through 2005 time frame.

Great.

Thank you very much.

Your next question is from Clark Westmont with Smith Barney.

Hi, guys.

George, I'm not sure if you answered about the SoC's ramp for desktop.

Okay.

Well, I think as we've commented before, you know, I think at the end of last year, we completed that ramp with Samsung --so that's, you know, a steady state now.

We are engaged with WD.

I think, again, we are, as you know, we're pretty careful about saying specific things about customers.

We know that WD has our SoC and it's certainly in their plans but as far as what they've done with it so far, they really, you need to as far as where they have it in their production plans and so forth really needs to be posed to them as opposed to us..

You say that --

It's there, but I think we don't want to, you know, necessarily get in front of that.

All right.

I'm sorry, that last part, it's there?

Oh, yeah.

No, no, it's ready -- It's imminent let's put it this way.

It's imminent.

So you are ready, just a matter of their timetable?

It's their call.

Okay, fair enough.

Thanks.

Okay.

Your next question is from Srinivas Pajjuri with Merrill Lynch.

I have a couple questions.

The [inaudible] on the customer side.

When do you think we'll see some penetration on the consumer pc?

And then you talked about the [inaudible] on the infrastructure side.

Given that there's no real need for the speed and also the product differentially seems like it's really significant, I mean, what do you think will drive gigi penetration on switching sides?

Okay.

The gigi for the consumer, we are actually already seeing gigi's for the laptop believe it or not.

This is the last thing that I even--I ever imagined gigi's would go into, gigi's for the laptop.

But the pricing on the gigi is such that it clears a huge depreciation of access to the companies are selling those products into the market.

You can argue that you may not need gigi for the laptop but if the price difference is so small and it [inaudible].

Similar to usb, usb too are very much employed in every pc.

Do you use the usb, too?

No, probably not for most applications your printer probably cannot print any faster than usb 1speed.

However there will not be a single pc you can ship without usb 2 today.

So a similar thing will happen to usb, gigi will go to pretty much in every pc.

The consumer size will be happening later because they are more cost sensitive.

But already in the white box, the high performance white box consumers home machines comes with gigi already.

You can buy it actually today if you want to buy white box, mother board pc, comes with Gigabit already.

Comes with powerful graphics, basically full picture.

And on the switching side, I guess my question was, given that there's no real need for the speeds and also it seems like the price difference is still kind of significant.

What do you think will drive this growth there?

On the switching side, actually, it's even easier to justify a move to gigi.

The different costs actually are small enough that the costs of hiring the IT personnel, worrying about which machine may need gigi, which machine may not get gigi.

It's too much hassle to deal with, so on the Enterprise side it's very much we'll move to gigi.

If the small -- I mean, minute, small percentage of the cost of doing business, for the Enterprise.

Okay.

One last question for George.

Any update on IBM on the design window?

Well, again, I think we're number one, very excited about what that opportunity is going to present for us.

It's a process that the new company, which is called HGST is going through, and, you know, our feeling is that when, you know, they make their decisions, which we hope they will come to some fairly soon, you know, we expect that to be very, very beneficial for us.

Nothing to report definitive yet.

Thank you.

Mm-hmm.

Ladies and gentlemen, we have reached the end of the allotted time for questions and answers.

Gentlemen, are there any closing remarks?

Well, there are no further questions?

This completes our Q2 fiscal 2004 conference call.

I would like to thank you, every one of you for joining us in this call.

And looking forward to updating you for the next quarter.

Thanks.

Thank you.

Thank you for participating in Marvell Technology Group ltd second quarter 2004 earnings conference call.