邁威爾科技 (MRVL) 2004 Q4 法說會逐字稿

Ladies and gentlemen, thank you for standing by.

Welcome to the Marvell Technology Group Ltd. fourth-quarter 2004 earnings conference call.

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

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

This conference call is being recorded today on Thursday, February 26, 2004.

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

Thank you, sir.

You may begin.

Thank you, Cody (ph).

Welcome everyone to our fourth-quarter fiscal year 2004 conference call.

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

Today we are pleased to announce the results of another strong year and fourth quarter for Marvell.

Once again, we generated record revenues.

In fact, the fourth quarter was our 25th consecutive quarter where we generated record revenue growth.

This (indiscernible) 2004 revenue grew 62 percent from the prior year.

Our 13th sequential increase in our quarterly revenue represents our ninth consecutive quarter that our sequential revenue growth has been greater than 10 percent.

We, of course, are also focused on our bottom line profit.

During the year, we generated GAAP net income every quarter and now have reported pro forma net income every quarter for the last five years.

I am very proud to look back at such accomplishments.

These results reflect the hard work and dedication of our employees, the strong positioning of our world-class technologies, and the support of our customers and partners.

Although it is very satisfying to look back at such achievements, our focus and excitement is clearly on the future, as we continue to drive Marvell for long-term growth.

We have made great progress since our IPO 3.5 years ago.

We have consistently grown our revenues, improved our profitability and strengthened our balance sheet.

Marvell is now positioned as a recognized leader in the growing number of growing high-volume markets which we believe (ph) position (technical difficulty) for substantial growth and strong financial performance for many years to come.

For this fiscal year, we are forecasting to end the year with over $1 billion in revenue, a significant milestone for a Company that will have been publicly traded for just four years.

Based upon this significant growth to date and the continued forecasted growth, we believe that now is an appropriate time to announce that our Board of Directors has approved a 2-for-1 split of our stock.

Once ratified and approved by our shareholders at our annual general meeting, we believe the stock split will offer our investors enhanced liquidity of our shares and make them more attractive to a broader rage of investors.

I will elaborate more about our current growth drivers and business progress, but first I will have George give our Safe Harbor statement and provide more insight into our Q4 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 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 related to the pace of our business as we completed our fiscal year 2004, and the impact of 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; 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, recent 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 Q4 financials.

Marvell reports net income or loss in basic and diluted net income or loss per share in accordance with GAAP, and additionally on a non-GAAP basis prefer to as pro forma.

Marvell's management believes 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 operations.

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

Today we reported that net revenue for the fourth quarter of fiscal 2004 was a record 243.3 million, an increase of 61 percent over the 150.8 million reported for the comparable quarter in fiscal 2003 and a sequential increase of 13 percent from the third 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 41.3 million, or 29 cents per share diluted, for the fourth quarter of fiscal 2004, compared with pro forma net income of 20.3 million, or 16 cents per share diluted, for the fourth quarter of fiscal 2003.

Shares (ph) using the computing pro forma earnings per share diluted for the fourth quarter of fiscal 2004 increased to 144.1 million, as compared to 129.3 million shares for the fourth quarter of fiscal 2003.

As we have for each quarter of fiscal '04, we continued to increase our net income under generally accepted accounting principles, or GAAP.

Our fourth quarter fiscal 2004 GAAP net income was 19.8 million, or 14 cents per share diluted, compared with a net loss under GAAP of 24.2 million, or 20 cents per share diluted, for the fourth quarter of fiscal 2003.

We have provided on our website in the investors' section at www.Marvell.com a reconciliation of GAAP net income or loss to pro forma net income for the quarter reported today, plus the prior 8 quarters.

Next I would like to make some additional comments on our Q4 results.

Our Q4 revenue of 243.3 million was another quarterly record for the Company.

Additionally, the 13 percent increase in revenue from Q3 to Q4 compares favorably to our Q4 guidance of a 10 to 12 percent sequential increase in quarterly revenue.

During Q4, our storage and communication businesses performed well, resulting in increasing revenue for both businesses.

Additionally, our performance in rechannels, SOCs, Serial ATA, as well as Gigabit Ethernet in both client and infrastructure applications was excellent, and all contributed to our (ph) sequential revenue growth.

From an end market perspective, was saw stronger than anticipated contribution from mobile storage, as well as 802.11b and g Wireless.

For the quarter, storage products contributed approximately 50 percent of total revenue, with communication products representing the balance.

Q4 gross margin at 52.2 percent was consistent with our guidance.

Q4 product mix, which is the largest determiner of gross margin percentage, was driven primarily by our increasing revenue into the wireless market and PC desktop applications.

Our end markets remained very competitive and during Q4 we continued with our cost reduction programs to achieve our targeted gross margins.

While we continue to invest increasing amounts in the development of new products that will provide us opportunities for growth over the next several years, we continue to experience the operating leverage from the increasing revenue contribution from products developed over the past several years.

In Q4, our pro forma operating expense as a percentage of revenue declined to 33.6 percent, which represents a 180 basis point decline from Q3 and 60 basis points favorable to our original Q4 guidance of 120 basis point decline.

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

Our continued double-digit sequential revenue growth, healthy gross margin percentage contribution and decreasing pro forma operating expenses as a percentage of revenue resulted in a 90 basis point increase in Q4 pro forma operating income percentage, from 17.7 percent to 18.6 percent.

The result of this continuous improvement has allowed us to achieve our long-term model of 18 percent for pro forma operating income.

Shares used in computing pro forma net income per share for the fourth quarter are increased to 144.1 million, compared with 142.4 million for the third quarter of fiscal 2004.

The increase in shares was driven by the use of a higher average stock price for Marvell shares in the treasury stock method calculation.

Exiting fiscal '04, our balance sheet remains very strong.

During Q4, we generated approximately 47 million in cash.

Also during the quarter, we made a onetime payment of 61 million to complete our purchase of our new campus facility.

For fiscal '04, we generated over 185 million in cash and exited the year with cash and short-term investments of 386 million.

DSOs for Q4 remained the same as Q3 at 50 days.

Our DSOs continue to remain in the range of our guidance of high 40s to low 50 days, and are likely to remain at the high end of our range during these periods of rapid revenue growth.

During Q4, we increased our inventory by approximately $15 million to support the current and projected growth of our business.

Our days of inventory increased from 69 days to 71 days.

The target range for days of inventory has been 65 to 70, but as the second half of fiscal '04 progressed, it became apparent that fab capacity was tightening and fab leadtimes were increasing.

As a result, we needed to adjust our plans.

Beginning in late Q3 and continuing into Q4, we began building additional buffer inventory, which will most likely raise our days of inventory.

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

Thanks, George.

As we start fiscal 2005, we are very excited that this will be the year that Marvell breaks through the billion dollar revenue mark.

Because of the strength of our technology and the large size of the markets we currently address, there continues to be very sizeable growth opportunities of our existing portfolio of products in the markets we serve today.

Additionally, as we have been discussing, we have been investing a great deal of R&D on new products and technologies to go after additional large markets.

Our strategy continues to be to leverage our core analog, mixed-signal and DSP technologies into very broad and large market opportunities.

Our recent introduction of our Power Management solution is just the beginning of such introductions.

We are also heavily investing in additional new innovative off-the-shelf, general (indiscernible) and linear components that we will also be introducing for broad markets.

Such new products will not only greatly expand our served markets with increased silicon content, but will now also expand our reach into both new vertical and horizontal markets, and take our products to thousands of customers.

I would now like to discuss the current positioning and growth prospects of our existing broad portfolio of products.

First, we are very excited about the strong growth opportunities still ahead for us with our storage electronic products.

Since we introduced our initial storage products a few years back, we have been able to grow our revenues at a very fast rate by both gaining market share and by leading the rapid adoption of system on a chip technologies.

Today, we are the only supplier of system on a chip, or SOC, across all classes of storage products, including the entry-level enterprise, high-performance desktops, low-power 2.5 inch mobile drive and the emerging small (indiscernible) drive.

This gives us a very unique and strong competitive advantage.

The fact that we can provide our SOCs across all segments allows our customers to leverage their firmware and software investment across all their product offerings.

This significantly drives down the cost and accelerates their ability to introduce new products into different market segments.

We continue to focus on driving strong growth from our storage electronic in many different market segments.

Even with all of our success to date, when you consider all the different silicon opportunities available to us, we still have only around 20 percent share of the targeted markets.

This shows the huge growth opportunities that still lie ahead of us.

In addition to the huge success of our SOC and rechannels, we see strong adoptions and large opportunities for our pre-amplifiers, motor controllers and other products for storage applications.

The adoptions of these storage applications in the consumer market is presenting Marvell with tremendous growth opportunities.

The technology (indiscernible) demands of these new applications are extremely challenging.

We know this because we have been investing heavily for many years, addressing the very specific issues such as extreme low power and higher reliability.

As a result of this tremendous effort, today all 1.8 inch type drives in production use our silicon, and we are designed into every single 0.85 inch drives in design today.

We offer these products with complete software and the capability to handle all of the different consumer interfaces that are found in such kind of devices.

We are very excited about being the clear leader in this emerging market.

Unit volume potential for these (indiscernible) storage applications is huge.

The .85 inch drives are clearly positioned to replace flash memory across many consumer electronic devices.

The advantages of these drives compared to flash memory are clear.

At the same cost as flash, these small drives are able to provide 10 times the storage capacity, consume low power and ultra-high throughput basically in the same small form factor (ph).

Why would anyone choose a device that only has 256 MB of memory when at the same cost, you can have greater than 2 GB of memory and have longer battery life and better performance at the same time?

We expect that the wise adoption of such drives in consumer electronics is going to result in unit volumes of these drives to exceed the (indiscernible) of desktop segments over the next several years.

Already these emerging drives are experiencing strong adoption, as evidenced by the success of the Apple iPOD in the 1.8 inch form factor.

And now the 0.85 inch (indiscernible) available, you will see much broader adoption in such high-volume applications, such as cell phones and digital cameras.

The consumer electronics market also presents very exciting opportunities for our wireless LAN technology.

Not only are the unit volume potentials huge, but the technological demands of such devices plays right into our strengths.

We have made years of investments in addressing these specific technological challenges.

As a result, we are driving widespread adoption of wireless LAN into consumer electronic devices by offering the highest performance, lowest power and smallest (indiscernible) of products, allowing for the most cost-effective designs.

We are winning numerous designs with large consumers, electronic OEMs, for applications such as (indiscernible) handset, gaming devices, PDAs and emerging home entertainment multimedia client solutions.

Our solutions enable our customers to easily integrate wireless LAN into their existing consumer electronic devices, because we are offering a complete solution set which includes a rapid (ph) design operating (ph) software and complete support for all of the variety of interfaces used in consumer devices today.

Additionally, and very importantly, our solutions come with a fully integrated, very powerful application processor.

This application processor offloads (ph) the system CPUs from the burden of processing wireless LAN packet and even allows for additional bandwidth for other applications such as MP3 decoding.

With such a uniquely positioned offering, we are having tremendous traction across a variety of consumer electronic platforms.

Our customers consistently tell us that we offer the lowest power in any operating condition.

We are also seeing strong adoption and revenue ramp of our new 802.11g solutions in the more established wireless type markets.

Q4 marked our first full quarter of 802.11g shipments, and we experiencing very strong initial success by building upon our strong installed base of 802.11b designs with the leading suppliers of (indiscernible), router and access point products.

We are excited about the growth of such applications as wireless LAN continues its rapid adoption.

Moving over to Gigabit, Marvell announced last week and demonstrated at Intel's Developer Forum the world's smallest PCI-Express (ph) Gigabit controller, which we call the Yukon-EC.

Its industry-leading features include our virtual cable tester technology, the advanced power management including highly efficient (indiscernible) LAN support, along with a very comprehensive software suite.

We again have worked closely with Intel to ensure full interoperability to the new PCI Express standard and are fully aligned with Intel's PC Express strategy.

With the Yukon EC, we are very excited to be able to continue to strengthen and build upon our existing partnership with Intel.

Design activity across all segments of the PC market is very strong and we are pleased to have already announced design wins as Asustek, ECS, Gigabyte and Micro Star, who are the four leading motherboard makers in Taiwan.

We are looking forward to the industry board broad adoption of PCI Express and to continue upon the strong success we have experienced with all the PCI devices.

Not only do we believe that we strengthen our market position even further with PCI Express, but we expect a strong acceleration of Gigabit as a replacement for (indiscernible) Internet as PCI Express is widely adopted.

As PCs are rearchitected around PCI Express, the Internet controller needs to be changed due to the changes in the interface.

Many designers will use these changes in the Internet controller as an opportunity to upgrade to Gigabit to Power g.

Also the improved system performance utilized in PCI Express further allows our Yukon Gigabit advantages to really shine.

The adoptions of Gigabit in infrastructure switching also continues to gain momentum and we are seeing a nice ramp of our Prestera switches.

We are now shipping Prestera in high volumes across all segments.

With our proven switching technology, Prestera leads the industry with the most advanced features, which have allowed us to have great success in the high-end markets such as (indiscernible) switches with our Prestera MX family and high (indiscernible) sophisticated enterprise class switching with our Prestera EX family.

We now have customers in volume production with high-end chassis-based systems and high (indiscernible) fully managed enterprise switches.

During the quarter, we also commenced volume shipments of our entry-level Prestera DX switches into unmanaged layer two and layer three valued-based systems.

Those are just a few highlights of some of the large growth opportunities and design activities currently in front of us with our established portfolio of products.

I would like to now spend a little bit more time on our recently announced entrance into the power management market.

Not only is the power management market a huge market opportunity for Marvell, but the market is also estimated to grow at a very healthy compounded growth rate of 18 percent.

The power management market was a $3 billion market in 2002, and is estimated to grow to approximately $6 billion in 2005.

We are very excited about the timing of our entry into this large market and our unique positioning to drive the adoption of our DSP-based power management solution as a replacement to conventional analog-based solutions.

Pure analog power management solutions are struggling with many technology code issues today, and these solutions are having trouble keeping up with the challenges created by the continuing advancements of technology.

We believe our DSP approach is the no-brainer answer to these challenges.

There are many examples of (indiscernible) technology transitions from analog to digital solutions.

Clear examples are the transition from analog modems to digital modems, be it DSL or cable; the transition from analog cell phones to digital cell phones; and the current transition underway from analog to digital TV.

These transitions offer tremendous opportunities for companies that possess the core technology to drive such transitions.

This is exactly what we did with the rechannel technology within storage electronics.

When we introduced the world's most advanced DSP-based rechannel, the entire industry was struggling with the performance limitations of pure analog rechannels.

However, with the adoption of our DSP-based rechannels, our solutions quickly forced any supplier who wanted to continue and serve the market to follow our lead and also develop DSP-based solutions.

Well, here we are again, but this time the current market is even larger.

Analog power management solutions will be replaced by DSP solutions.

It just makes sense.

DSP solutions offer better performance, faster response time, smaller external components and improved reliability.

I predict that just as today you cannot find a pure analog rechannel, five years from today you will not see a pure analog power management solution.

With years of heavy investment in this technology, we are strongly positioned to lead this transition.

We have always worked closely with our customers throughout the design process.

This creates the opportunity to share the benefits of our new solutions and designs of our power management products alongside of our other products.

Additionally, you should also expect to find our initial penetration into laptops, PC motherboards and PDA devices.

And of course, we are also focused on continuing our penetration into cell phones, industrials, automotive and the off-the-shelf, general purpose distribution market.

We are very excited about the potential of our power management solutions; however, we are not stopping here.

We have identified the very large traditional analog market opportunities where we can also drive the adoptions of our DSP-based solutions.

We plan on introducing a number of other linear DSP-based solutions into markets that have traditionally been performed by standard analog devices.

We look forward to updating you shortly on our progress and details of such products and markets.

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

Thank you, Sehat.

For the last several years, we have included as part of our Q4 call guidance for the new fiscal year as specifically for Q1.

Additionally, on this call, we will be providing an update of the long-term business model.

I will begin with the new business model.

Approximately two years ago, we provided to investors our long-term business model that we anticipated achieving over the next six to eight quarters.

In review, that model assumed 52 percent gross margin, 34 percent pro forma operating expenses, 18 percent pro forma operating income and a 12 percent tax provision rate.

The strategic assumptions used in anticipation of achieving a long-term model were, one, use our superior analogue mixed-signal DSP technology to create products that would put us in a leadership position in our target markets.

Two, focus on markets with large TAMs to maximize the rate of quarterly revenue growth.

Three, use our design expertise to create the smallest die sizes at a given process geometry for low cost; and four, continue with our investment in R&D to ensure a flow of products that position the Company for long-term growth.

During the last two years we have made steady progress towards achieving our model.

And over this time period, our annual revenue has grown from 289 million in fiscal '02 to the 820 million we reported today for fiscal '04.

Quarterly pro forma operating income has increased from approximately 2 percent to the 18.6 percent reported for Q4.

Marvell is growing into a much lager semiconductor company and we anticipate achieving the one billion revenue level for fiscal '05 and growing our total worldwide headcount to 2000 employees.

It is now a time to establish a new long-term model that reflects the change.

Our belief is that the strategic objectives that I just outlined will continue into the future and provide us the opportunity to continue increasing our pro forma operating income percentage.

A new long-term business model is a gross margin percentage of 51 to 52 percent, pro forma operating expenses of 28 to 30 percent, and pro forma operating income of 21 to 23 percent.

Our tax provision rate, beginning with fiscal '05, will be 10 percent.

We anticipate achieving this model in four to six quarters, but should again see steady progress towards the model beginning with Q1 fiscal '05.

Now turning to fiscal '05.

We continue to develop new products and target new markets to expand our TAM.

Entering into fiscal '05, we are very excited about the adoption of our products based on our analog mixed-signal DSP technology across a broad array of consumer platforms and our recent entry into the power management market utilizing our linear DSP technology.

These opportunities should certainly increase our TAM.

Additionally, our position in our established markets remains very strong and the technology changes that have been underway for some time continue.

Given this strong position, we believe that achieving the milestone up a billion dollars in revenue for fiscal '05 will happen.

For fiscal '05, our revenue should range between 1.125 billion to 1.175 billion.

At the midpoint of this guidance, our annual revenue growth from fiscal '04 to fiscal '05 would be 40 percent.

Gross margin percentage for fiscal '05 should range between 51 to 52 percent and pro forma operating expenses should range between 31 to 33 percent.

Now moving to Q1.

Our business momentum in Q4 was quite strong, as evidenced by our Q4 results.

Entering Q1, which is normally a seasonally slower period, we are encouraged by the continued ordering patterns from our customers, and consequently feel that seasonality will have no significant impact to our Q1 revenue.

The continued adoption of our technology in new product offerings from our customers and our continued market share gains against our competitors positions us for revenue growth in Q1.

We have targeted Q1 fiscal '05 revenue to increase to approximately 267 million, which is a 10 percent growth from Q4.

Our product mix of revenue in Q1 is expected to be similar to Q4, and therefore Q1 gross margin percentage should be consistent with Q4 plus or minus 25 basis points.

As I mentioned in the long-term model commentary, we expect to continue to express operating leverage beginning with Q1 fiscal '05.

While operating expenses in absolute dollars will increase to support our future growth, they will decline by 28 (ph) to 30 basis points from Q4.

Our balance sheet (indiscernible) should be consistent with the last several quarters and shares used in computing pro forma net income should increase to approximately 146 million.

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

Thank you, George.

That completes our commentary.

Cody, would you please poll for questions.

(OPERATOR INSTRUCTIONS) Ambrish Srivistava with Harris Nesbitt.

Just a couple of housekeeping questions and then onto the business.

If you look at the balance sheet, accounts payable was up pretty significantly, and also could you comment on the deferred income line?

It was up last quarter and then down this quarter.

How should we be looking at it?

And then also the cash flow from ops, D&A and CAPEX?

And then moving onto the business, if you look at the storage business, did I hear you guys correctly; was it 51 percent?

And if so, this business has been growing double-digit last couple of quarters.

What was the reason for the deceleration despite the fact that Western Dig should be ramping?

If you could please comment on those, thanks.

Why don't we do the second one first because you have the wrong number there.

We said basically that storage represented mid 50 percent, which is consistent with where it's been over the last several quarters.

And both storage and communications had very strong Q4s.

So back to your questions on the balance sheet.

We have been in a very aggressive ramp mode of our production in our commitments to our fab partners, and a lot of that -- as well as our back-end partners.

So a lot of that is reflected currently in our accounts payable.

And then under the deferred revenue line basically, we had some extraordinary items in there in Q3, which we weren't able to recognize the revenue, hadn't met all the tests to recognize the revenue, and were able to accomplish that in the Q4 time period.

So now deferred revenue is back pretty much to where I think it should be, which is going to be 10 to 12 million on a going-forward basis. 12 to 14 million going forward, I'm sorry.

Then what percent of sales was Intel, and could you comment on any changes on the assumptions on the Intel relationship?

Well, we're pleased to report we have three 10 percent customers.

We never give the specific percentages.

But the three 10 percent customers were Intel, Western Digital and Samsung.

I think we have already made some comments about the Intel relationship, but maybe Sehat, would you like to (multiple speakers) a little bit?

Yes, we mentioned I said with the Yukon-EC announcement, we are very, very excited, because it is Yukon-EC actually strengthened our partnerships with Intel.

It is in very good shape.

Jim Liang with Pacific Growth Equities.

Thank you.

One question for Sehat.

On the power management front, there is one end of the spectrum of extending battery life or portable devices, and the other being managing heat dissipation in processors.

Can give us a little more color as far as which areas would you see traction first?

We are actually addressing all areas.

The products that we have introduced in the last few months addressing for the battery management side first.

But in the next -- I guess in the next several quarters, you will see that additional products addressing higher and higher (indiscernible) markets will be introduced, including for managing very, very high current (ph) CPU types of applications.

Which (indiscernible) are we asking for because the challenge for CPU probably is even more challenging as the CPU goes to 90 nanometers and even 65 nanometers (ph) in the next several years.

So the requirement -- this is where the DSP solutions will shine, in those kind of applications.

Seogju Lee with Goldman Sachs.

Great quarter.

Just wanted to get a little bit of color in terms of Q1 guidance from a market perspective and also for the full year, if you could talk about how you expect the various segments to ramp and also when we might expect initial revenues from the power management parts.

Thanks.

I think in the commentary we alluded to a number of the areas that are doing well and that our business is really firing on all cylinders as we exit '04.

And then you will overlay in the consumer and power management opportunities that will begin to have some effect during this year.

And I think those kind of give you a flavor for how things should be driven.

As we indicated in our announcement of power management, we already are in production with power management, and as with all new products, the further in time you go, the larger the quarterly revenue becomes.

So this is the beginning and we are expecting it to grow sequentially every quarter going forward from here.

Arnab Chanda with Lehman Brothers.

A couple of questions.

Help me understand this a little bit.

It seems like some of the areas that are ramping right now are more in things like say pressure (ph) on the infrastructure as well as obviously power management, which would appear to be higher gross margin than your historical business.

If you could explain that a little bit, why it seems like you're sort of lowering a range of gross margin a little bit.

What part of this is (indiscernible) mix, what's happening with pricing, et cetera?

Thank you.

The growth is across the board, as George said.

The power management -- you are correct -- will have higher gross margin than the other products, probably significantly higher gross margin.

But because the (indiscernible) is just in the early ramps of the power management, the impact for the gross margin enhancement will not be seen for maybe a year or two (ph).

So in the meantime, we are still in the competitive markets, and we also where we mentioned in the call that the gross margin is within our long-term gross margin to begin with (multiple speakers).

I think if you look at the commentary that I made relative to Q4 and the strength -- or actually more strength than we had originally anticipated coming out of the wireless LAN area and coming out of desktop application, our margins, as our revenue is growing and we've gotten bigger positions in those markets, it has come down a little bit from its maybe peak margins.

I think a hint at your question is if you look at the guidance now for Q1, we said margins aren't going to -- are not likely to go down.

So I think that is reflecting the positive contribution from some of the higher margin pieces of our business coming into play.

With the market segment we are serving, our margin is actually healthier than our competitors.

The key reason is because of our efficient design and manufacturing capability.

Cody Acree with Legg Mason.

Weili, maybe you can follow up on the Gigabit Ethernet side.

Can you talk a little bit about -- I know you won't give too specific -- but a little bit of a detail as to how the revenue is split up between infrastructure and desktop, and what kind of a ramp you see happening now with Prestera.

First of all, I think the split detail, I think, we don't disclose.

As far as the Prestera ramp, it has been very healthy.

We have over a dozen customers shipping (indiscernible) with our Prestera based solution, all the way from high-end to very low-end, unmanaged systems.

We had told, I think, in previous commentary that we were anticipating six by the end of January; we actually are having (multiple speakers).

We actually had a dozen starting out here, so it's very, very positive.

Can you maybe give some geographic discussions or at least maybe some breakdown of where you are stronger?

You talked about some higher-end systems, maybe some lower-end systems.

Is there any concentration one way or the other you can give us?

Actually, across the globe.

And the high-ends in Asia and U.S., same as low end.

I think there's been four people who have announced.

Nortel which has many, many offerings.

Wawai (ph) in Asia and Terasis (ph) and then of course Dell, which we were very excited about around the CES show.

Dell has made a decision to go 100 percent of all of their power connect offerings utilizing our Prestera.

You mentioned that you were adding a little bit of inventory to manage what's been some level of constraints out on the manufacturing side.

How is that looking going forward?

What's the efforts to continue to secure more production, especially with this kind of a ramp?

What's happening with delivery leadtimes and is there any changes in prices?

A lot of questions there.

I'm sure it's the hottest (ph), but let me just address the leadtimes one and then Sehat may certainly have some other thoughts.

The leadtimes have gone out beginning during the fourth quarter, and the reason that we are actually building a little additional -- but we always have some buffer inventory because we have been in a mode for well over two years now of seeing continued increasing demand for our products, so we want to make sure that we are always any position to respond positively and satisfy that demand.

But as the leadtime through the fab and then the back end is extending, it makes it difficult to accurately predict how that will flow into what the customer demand is.

Again, we don't want to put ourselves in a position where we can't respond, so we are being even a little more aggressive in making sure we have adequate inventory to service that demand.

I think that's really the logic behind what we are doing there.

Especially for some customers that are just ramping up, we do not want for a minute to be the one that caused them not to be able to ramp up.

If we look at power management, we will build inventories because again, (indiscernible) an early ramp up of our production, we do not want to have any glitch at all in our production.

Cody was also asking about the capacity.

The capacity -- the fabs is (indiscernible) the fabs are all full.

The beauty is we have been working with our foundries for the last several years very closely and then when I say working (ph) meeting, we been partnering, we have been a good partner to them in the bad time when the fab was empty and when the fab is full, they remember that we have been always a good partner to them and we always will get capacity that we need within reasonable forecasts obviously.

So we are not seen any capacity -- inability to get the capacity we need.

David Wu was Wedbush Morgan Securities.

David Wu.

Excellent quarter.

Can you help me on two things?

Number one is when I was at IDF (ph), I guess the Intel silicon was shown.

And yet you have good growth in the client side -- (indiscernible) client side in the past quarter.

Both your Yukon and Intel grew.

And secondly, you implied that your relationship at Intel has gotten better.

Does that mean that you'll serve part of (ph) the future PCI express requirements and internal activity would only serve a fraction of that?

And the final question is George, if you have a 10 percent sequential growth in Q1, mathematically to get to 40 percent, some quarters will not be double-digit.

Which would the quarters be?

I think -- as we have in the past, David, we are always very excited about our growth prospects.

But when you try to look out a whole year, it's very difficult to predict four quarters exactly the way they are supposed to be.

Our belief is that we have a lot of momentum and at our revenue level above a billion dollars, if we are able to grow 40 percent year-over-year, there aren't very many companies that are able to say that.

So I think let's take the first quarter at 10 and then we will come back in three months and tell you about Q2.

The Intel piece, the 90 nanometer, we are aware, I think, that it is a very powerful silicone and this is mainly focusing on very high-end server platform.

That's very good for Intel.

As far as the partnership, as we said, it's been very healthy and moving forward, we will continue to partner together.

The detail, I think, at the end of day, we always wait for customers to introduce their products and we acknowledge that (indiscernible).

As far as our Yukon story, we have been very successful addressing (indiscernible) the desktop program.

As you can see we recently announced the design wins with major players in Taiwan, the Asustek, the ECS, Gigabyte and Micro Star.

We are also doing very well in terms of our own silicon.

We have expanded that.

When we first came out with Yukon, we just -- (indiscernible) part of the lead group.

And now we have -- within about basically a year we've gotten everyone, so I think that's (multiple speakers).

That's a very small footprint and very efficient solution; we believe this is going to be very critical for high-volume desktop platform.

(OPERATOR INSTRUCTIONS) Jeremy Bunting with Thomas Weisel Partners.

This is Rubin Roy (ph) calling in for Jeremy.

I was wondering first, can you give us the relative mix in trends in 11b versus the 802.11b/g chipset that you started shipping last quarter?

I don't think we have the numbers with us.

We started with b and now we are in the process of seeing g be more widely adopted.

Right.

We have very strong adoption of our b's, and then we are building upon our relationship with our b customers to adopt the g solutions.

Some of the customers will continue to buy the b solutions for certain market segments; other customers will slowly migrate to g.

Some customers (ph) might quickly migrate to g completely.

So it's hard to say what's the percentage or even predict what the percentage.

Either way, whether the g will be more percentage or less percentage, it doesn't matter to us because our solution -- both sides are the most (indiscernible) in the market.

So we will be competitive.

Charlie Glavin with ThinkEquity Partners.

Sehat, if you could take a look maybe at a different way rather than by end market.

It seems that both with the power management along with the wireless LAN coupled with your smaller drives that there seems to be a bigger shift by Marvell to getting more dollar content along several different designs within the same box, as evidenced with your design with Vixel and Emulex at first.

Is there any sort of metric that you can give us as far as what you're looking as far as the typical dollar per box or the type of designs, multiple designs you're looking at per box, and is that a right way of thinking on that?

Not just in terms of taking more market share or the end market growth?

To tell you the truth, I never look at what the dollar per box.

The power management market is completely -- it's a complete different market segment than where we have served (indiscernible) so far.

This market is the biggest market that we ever served to date, as is evident by the projected growth to be a $6 billion (ph) market.

Now of course, the market is very wide.

There are a lot of different solutions.

And this is the reason why we are coming out with the DSP solutions -- they are scalable.

The key is to make the solutions to be scalable so that we can go to many different markets very quickly as we win each top (ph) market segment in power management.

There will be existing markets there where we will also (ph) (indiscernible) power management, so that part will grow our dollar per boxes.

But is not the reason why we built those new markets -- I mean, with the power management.

We want to serve even bigger market segments, like (indiscernible) cell phones, which we have practically zero market shares in the past, so we want to go in -- we (indiscernible) in the call, we mentioned that we are addressing wireless LAN for cell phones, but we want (indiscernible) to address power management for cell phones.

It is a huge market opportunity.

Aalok Shah with Pacific Crest.

Maybe a couple quick questions maybe on -- a little bit on some of those devices that you (indiscernible) had mentioned already in the power management (indiscernible) that you're starting to see it in the laptops and PC motherboards (indiscernible).

Are there devices that you're working on maybe that integrate some of your other solutions.

Can you least talk conceptually about that and maybe give us an idea of where we might be seeing those types of (indiscernible) in the future?

Power management is actually a stand-alone device.

It is not integrated with any other functions.

The purpose of our power management device is to serve as wide a market as possible.

The idea is to -- any devices that you need that use advanced geometry, like the .13 micron or below, will need a more advanced power management.

What we want to do is we want everybody who built products using those advanced (indiscernible) to use our power management device, which is based on DSP, to give them a better system overall performance.

The laptop -- there was a question about laptop.

So the laptop is the target market.

In the laptop, typically you'll see about a dozen to -- maybe 12 to 15 or 16 power management devices in each laptop.

We of course will focus heavily in this market segment.

Srinivas Pajjuri with Merrill Lynch.

A quick question for George.

Could you talk about what's driving the tax rate lower going forward?

We've been using the 12 percent, I believe, for a little more than two years.

Once we had integrated the Galileo Technology acquisition into our structure, we have now looked at and have had tests by different regulatory authorities throughout the world and come through very, very successfully with those.

So if you go and look at our balance sheet, you see a very significant number for taxes payable, which has been the difference between our book rate and our filing rate.

So based on the positive responses that we've been getting, we now feel that it's more appropriate to use that 10 percent rate to more closely approximate -- it's still higher than what our filing rate is, but more closely approximate that rate, and then just puts a reasonable degree of risk between what the book rate is and what the filing rate is.

Karl Motey with Wachovia Securities.

Most of my questions have been answered.

Sehat, one clarification.

In your opening remarks, you mentioned on the storage SOCs that you have penetrated only 20 percent of your targeted end markets.

Do I have that number correctly or did you mean -- was there another number that you used?

What I meant 20 percent is the overall storage market, not specifically in the SOCs.

In the SOCs is a bigger number -- I don't have the number with me.

But if we're talking about the overall market including all the different components (indiscernible) in the storage.

Right.

So it's well beyond just the rechannel and SOC, Carl.

It includes the pre-amps, motor controllers.

It's a multi, multibillion dollar market when you add all those things in.

And we have the technology and product to serve those markets.

Max Schultz (ph) with Credit Suisse.

Hi, it's Jeff (indiscernible) for Max.

Can you just give some color on your non-Western data storage business?

What type of growth did you see, what types of trends did you see?

And then also, you have touched some on the consumer opportunity, but when does that ramp?

I think again in the commentary that I made relative to the Q4 revenue, I said that we experienced growth that would be in the storage side of our business for rechannels, for SOCs and for serial AT As.

So in other words, all of our different products within storage electronics, all did well in the fourth quarter.

Probably the end market segment we said that was the strongest for us and was a little more than what we had maybe originally expected going into the quarter is no surprise to anybody at this point, is the mobile, since there's been so much commentary about how well people did with that.

So that was the end market.

But overall, all of our products are doing extremely well.

I think realistically on the consumer, it's going to be later this year before we would see anything that I would call material type of revenue from those drives. (multiple speakers) the 1.8 inch which we are already shipping today, but this is for 0.85 inch products.

So the 1.8 inch will continue to grow as the adoptions of like (indiscernible) devices or like type of devices ramps up.

But for the smaller compact (ph) for the .85 inch that will fit in the cell phones (indiscernible) type of packages will grow very rapidly.

My prediction is that in the next several years it could be bigger than the desktop market today.

So this is a huge, huge market opportunity for us.

We been working on this for close to four years.

So we been developing and solving all the problems that are intrinsically different than the other markets.

So we are very, very excited.

The application that is just mind-boggling is cell phones, MP3 players, game machines, portable game machines, PDAs, automotives -- it could be anything.

It is going to be measured in the hundreds of millions of units potentially.

Right.

Shah Wu (ph) with American Technology Research.

Great quarter.

Could you talk a little bit more about -- you made some comments about leadtimes and capacity tightening with your fab partners.

Any color on impact on your margins?

I think we have already said quite a bit about it.

Our margins remain very healthy and we continue to reduce our cost across for our products.

We don't see the current capacity utilization increasing.

The leadtime should have absolutely no impact on cost, but the tightening of the capacity and so forth is just where we are in the cycle, and we are able to deal with that, and we are constantly reducing our cost.

Allan Mishan with CIBC.

A couple quick ones on the guidance.

First, on the full year, is there any ramp down in sales to Intel on an Intel internal par coming in embedded in your guidance, either at the low end or the midpoint?

Second, is there any power revenue in that guidance?

And then with respect to the 10 percent for the first quarter, I would imagine Intel sales would probably beat down, flat at best, just because of the volumes for seasonality.

So what's making that up?

Allan, we are not going to get to the level of predicting each individual customer.

I think the client opportunity and the market share that we have with Intel is pretty well documented.

So I think you can look at the end number of units being shipped on a quarterly basis and determine how we are going to do there.

I don't think want to be any more specific on that, except that the demand is supposed to increase as you go through the year.

Answer to the second question is, yes, there is absolutely some power management included now in our estimate for the year, because we are already in production with the product and therefore we have revenue for it.

Relative to seasonality for Q1, yes, this is the seasonally more challenged quarter, but as we said, we don't expect -- while there is some impact, it's not meaningful, and certainly the Intel piece of our business may fall into that category.

Quinn Bolton (ph) with Oppenheimer.

Just a quick clarification and then a follow-up question.

On the Intel relationship, can you comment whether or not you're working on a PCI Express-based Gig E (ph) controller with Intel?

And then the following question, if you could just kind of talk about the desktop SOC opportunity -- is that ramp (ph) and sort of where are you in that ramp?

Thanks.

(indiscernible) beat on the Intel question for the next five years.

The desktop, as you are aware, we entered the business we started in enterprise, and then we moved to put our eye onto the mobile.

And (indiscernible) the reason why we are doing very well in mobile, because we put a lot of efforts in solving a lot of mobile problems, meaning power consumption, small form factor and things like that.

More recently, we are putting our eye on the desktop.

So we are clearly targeting desktop as our next opportunity for growth, because the desktop is clearly still the biggest market segment until the small compactor takes off.

So we are very hopeful that our device with the capability of the enterprise will be very, very useful for the desktop.

As the performance of the desktops are going up, there is a prediction that the new technology of vertical recordings coming into the desktop in the next couple of years, the speed requirements of the desktop may approach the enterprise speed requirements.

And this is where we are really, really strong at.

On the Intel one, again, we are not going to get into that level of detail.

I think the fact that we've done a number of chips now, probably approaching a dozen chips throughout the partnership that we've done with Intel.

So I think we will leave it at that.

We're very -- as we said in our commentary, we are excited about our relationship with them and it's very good and just stay tuned.

Probably the most we can say at this point.

Are there any closing remarks for today's conference call?

Thanks, Cody.

This completes our Q4 fiscal year 2004 conference call.

I would like to thank all of you for joining us and look forward to updating you for the next quarter.

Thank you.

Thank you.

This concludes today's conference call.

Thank you for anticipating.

At this time, you may disconnect your line.