英特爾 (INTC) 2005 Q1 法說會逐字稿

Good afternoon.

My name is Brian, and I will be your conference facilitator today.

At this time, I would like to welcome everyone to the Intel first-quarter 2005 earnings conference call.

All lines have been placed on mute to prevent any background noise.

After the speakers' remarks, there will be a question-and-answer period. (OPERATOR INSTRUCTIONS).

Mr. Doug Lusk, Director of Investor Relations, you may begin your conference.

Okay, thank you, and welcome to the Intel first-quarter earnings conference call.

Attending from Intel are CFO, Andy Bryant, and President and COO, Paul Otellini.

Before we begin, please bear with me a minute while I read our Safe Harbor language.

The first-quarter earnings report discusses Intel's business outlook and contains forward-looking statements.

These particular forward-looking statements and all other statements that may be made on this call that are not historical facts are subject to a number of risks and uncertainties, and actual results may differ materially.

Please refer to our press release for more information on the risk factors that could cause actual results to differ.

The specific forward-looking statements cover expectations for product mix and demand, revenue, gross margin, expenses, tax rate, interest and other income, capital spending, depreciation, and amortization of acquisition-related intangibles and costs.

These statements do not reflect the potential impact of any mergers, acquisitions, divestitures or other business combinations that may be completed after April 18, 2005.

Lastly, if during this call we use any non-GAAP financial measure as defined by the SEC and Reg.

G, you will find on our website on intc.com, the required reconciliation for the most directly comparable GAAP financial measure.

And with that, I'll turn it over to Andy.

Andy?

Thanks, Doug.

At time of year when business is typically subdued, we achieved good year-to-year growth and revenues led by both -- computing, flash memory, and chipsets.

Revenue operating income and net income each registered double-digit growth from a year ago.

Although revenue declined from the fourth quarter, profitability was higher.

Factories producing microprocessors and chipsets were full, which improved revenue and profits.

Gross margin for the quarter was much better than we expected, and we have raised our outlook for the year.

With the supply and logic products tight and the deployment of 65nm process bearing well, we have increased our annual budget for capital spending.

We can also report progress in building inventories.

The increased dividend payments and accelerated stock repurchases are returning record amounts of cash to stockholders.

And shares outstanding are at the lowest level in years.

Revenue for the first quarter was $9.4 billion just over the top of the range we forecast in the January earnings release and in the March update.

The first quarter was 14 weeks in contrast to our typical 13 weeks.

Although it is impossible to estimate with precision the revenue differential from the extra week, we believe the decline of 2% from the fourth quarter is in line with underlying seasonal patterns typical for this period.

Revenue has been down in the first quarter for each of the previous 5 years.

On a year-to-year basis, revenue was up nearly 17% from the first quarter of 2004.

A year-to-year comparison probably benefited from the extra week, and we do not expect to sustain this pace in the second quarter.

Unit shipments of total microprocessors were approximately flat for the fourth quarter, while unit shipments of chipsets, motherboards, flash memory and wired connectivity units were lower.

Unit shipments of wireless connectivity units were higher.

In January, Intel announced a reorganization of resources and people to make Intel's structure consistent with our strategy to focus on platform products.

As a new business unit by far the largest are the Mobility Group and the Digital Enterprise Group.

Each of which will also be reportable segments.

Our goal is to present financial results for these operating segments in the 10-Q for the first quarter to be filed in the next few weeks.

Today, I will focus on revenue of major product categories.

Total microprocessor revenue for the first quarter was $6.9 billion, down approximately 1% from the comparable number in the fourth quarter and up 12% from the first quarter of 2004.

In addition to revenue from the Intel Architecture business microprocessors, which we previously disclosed, this also includes revenue from other microprocessors such as embedded processors.

Revenue for flash memory products was $578 million.

This is a decline from the fourth quarter of 10% and an increase from the first quarter of 2004 of nearly 39%.

Revenue from chipsets, motherboards and other products was $2 billion, flat sequentially and up nearly 29% year-to-year.

Gross margin dollars of approximately 5.6 billion increased 4% sequentially and 15% year-to-year.

On a percentage basis, gross margin was 59.3%, over a point higher than our forecast range at the mid-quarter update.

Most of the improvement since that time came from lower unit costs, a richer product mix, and earlier product qualification.

The gross margin percentage is over 3 points higher than the fourth quarter.

The key drivers of this sequential progress were lower microprocessor unit costs, sales of inventory that had previously been written down, and the absence of a special year-end employee bonus.

In the year-to-year comparison, gross margin percentage dipped slightly lower than the first quarter of 2004 on a charge related to a legal settlement, reduced gross margin by about 2 points.

Spending on R&D, marketing and G&A was $2.5 billion at the lower end of our forecast range.

The higher number of employees and an extra work week accounted for much of the sequential spending increase of 4%.

Compared to the first quarter of 2004, spending grew at a slower rate than revenue and was also a smaller percentage of the revenue.

The number of employees rose during the quarter to 87,000 at the end of March, a year-to-year increase of 8%.

Operating profit, as a percentage of revenue at 32%, was higher than either the previous quarter or the first quarter of 2004.

Operating income was 3 billion, a sequential increase of 5% and a year-to-year increase of 22%.

Operating income a year ago included $162 million deraid (ph) to a legal settlement.

A legal charge accounted for 7 points of the 22% year-to-year increase.

The total in the first quarter for interest income, other income, and gains and losses on equity investment was $119 million, which is higher than our forecast of 100 million.

Within this category, the income statement, interest and other income was $115 million.

Fully diluted earnings per share, which include potential dilution attributable to employee stock options, was $0.34.

Basic earnings per share, which does not include potential dilution, was $0.35.

Diluted earnings per share of $0.34 is $0.01 or 3% better than the fourth quarter.

The year-to-year improvement is 31%; although, the first quarter of 2004 included the charge from a legal settlement, which amounted to $0.017 per share.

Earnings per share grew at a faster rate than net income, a function of the progress we have had in reducing share account with stock repurchases.

Average shares for calculating diluted earnings per share were 6.3 billion.

Basic shares outstanding were 6.2 billion, down 1% from the fourth quarter and 4% from a year ago.

Compared to the peak in the second quarter of 1998, basic shares are down approximately 8% or 550 million shares.

During the quarter, we repurchased 108 million shares at a cost of $2.5 billion.

The objective of the buyback program is to return cash and increase the return to stockholders.

On the balance sheet, inventories of $2.8 billion were higher than the fourth quarter by $187 million or about 7%.

This is an improvement from January, when we believed inventories were at levels that were less than optimal.

Raw materials were approximately flat, while the rest of the increase was split between work in process and finished goods.

On a product basis, most of the increase came from chipsets, wireless connectivity and flash memory.

Total cash investments comprised of cash, short-term investments, and fixed income trading assets ended the quarter at $15.8 billion, a decrease from the fourth quarter of $1 billion -- after stock repurchases of 2.5 billion, capital spending of 1.8 billion, and dividend payments of nearly $500 million.

As we turn now to the outlook for the second quarter, please keep in mind that the forecast data do not include the effect of any new acquisitions or divestitures that may be completed after April 18th.

I will use the mid-point of forecast ranges when making comparisons to specific periods.

We're planning for revenue in the second quarter to be between 8.6 and $9.2 billion.

The mid-point of this range will be down 6% in the first quarter.

While the midpoint of the range is down 6% compared to historical average declines for this period of 2%, we believe the change is consistent with historical patterns when the extra week in the first quarter is considered.

Compared to the second quarter of 2004, the mid-point of the forecast revenue range would be an increase of approximately 11%.

Our expectation for gross margin percentage in the second quarter is 56% plus or minus a couple points.

This is more then 3 points lower than the first quarter.

The decline is caused by several factors.

First, lower revenue in general leads to lower gross margins.

Second, the mix of products should also contribute to lower margins, as we increase the supply of chipsets.

And finally, with the ramp of new 65nm process technology, start-up costs should be approximately $100 million higher in the second quarter than in the first.

Spending -- R&D plus G&A should be approximately $2.6 billion.

Annual salary increases take effect at the beginning of the second quarter and drive spending a little higher.

There will also be a flat increase in project spending.

Depreciation should be $1.1 billion, plus or minus $100 million.

We expect amortization of acquisition-related intangibles for the second quarter to be approximately $35 million.

Our estimate for gains and losses on equity investment and interest and other income is a net gain of $70 million.

For the entire year, we have raised our outlook for gross margin percentage by about a point to account for the better-than-expected gross margin in the first quarter.

The revised forecast for the year is approximately 59% plus or minus a few points.

The mid-point will be more than a point higher than 2004.

We have also raised our forecast for capital spending for the year and are now targeting $5.6 billion plus or minus $200 million.

This increase reflects the demand-supply situation in the recent quarters, a slight increase in the outlay for equipment, the demand for 65nm process technology, and a variety of smaller projects.

The outlook for R&D spending is unchanged at approximately $5.2 billion.

The depreciation forecast remains $4.4 billion plus or minus 100 million.

The estimated tax rate for the second, third and fourth quarters is approximately 31%.

In summary, the first quarter was a solid start to the year.

While it is always appropriate to be concerned about the economic outlook, we feel we are on firm footing with the strength of our business model.

The economics of the model are inherent in the unique aspects of silicon technology, as predicted by Gordon Moore 40 years ago.

The business is generating consistently high levels of cash, as we move forward with new processes and facilities to make high volumes of advanced products at lower unit costs for an expanding array of uses and users.

With that, let me turn it over to Paul for comment on the business.

Thanks, Andy.

Revenue set a record for the first quarter of 17% versus last year.

Growth was led by strong demand for our mobility products including notebook prank (ph) microprocessors, wireless LAN complements, and application processors for handsets.

Along with good results, we also made significant progress in aligning the Company around platforms and launched new mobile, server and desktop platforms.

Turning to our geographies, revenue in the Asia-Pacific region was approximately flat sequentially and grew 34% year-over-year, as we continued to see good growth in emerging markets, such as China and South Asia which includes India.

Mobile processor demand continued to show strength, and chipsets sales exceeded our original expectations.

In Europe, revenue was down 8% sequentially but up 9% versus last year.

Sequential growth in mobile partially offset a decline in desktop units.

Our channel business was easily normal in Western Europe but set a new record in the Middle East, Turkey and Africa region.

In the Americas region, year-over-year PC consumption was up by a double-digit percentage with mobile form factors showing the strongest growth.

Our revenue was lower year-over-year due in part to the ongoing shift of PC production to Asia.

In Latin America, revenue was up sequentially, setting a new record and showing solid double-digit growth versus last year.

Japan had all-time record revenue, driven by strong domestic demand as well as notebook exports.

We also saw good flash demand in 2.5G and 3G handset designs and in the embedded market segment.

Our worldwide channel business grew slightly from the record levels set in the fourth quarter.

Processor unit sales in the channel were about flat overall, and we saw a very good demand for our cellular and handheld products.

Moving onto a discussion of our new platform groups, I would like to start with the Digital Enterprise Group, which had a number of significant product launches in the first quarter.

Yesterday, we announced that PC makers have begun selling systems based upon our first dual-core processor, the Pentium Processor Extreme Edition 840, which provides two processor cores and HT Technology -- allowing up to four software threads to run simultaneously.

We also announced plans to develop a dual-core Pentium D processor-based platform for mainstream PCs later in the second quarter.

We expect to ship millions of dual-core processors this year and rapidly ramp into higher volume in 2006 on 65nm's.

We will be providing more details on our dual-core plans at our analysts' meeting in a few weeks.

Also for the desktop, we introduced five new microprocessors that bring 64-bit capability to Intel-based desktop PCs.

The new Pentium 4 processors in the 600 sequence offer 64-bit computing capability, HT Technology, 2 MB cache, and enhanced SpeedStep Technology.

We also announced plans to make 64-bit capability available throughout our desktop processor line-up later in 2005, including the Celeron D processors.

We introduced a new platform for mid-tier servers that includes the first 64-bit Xeon MP processors, allowing us to offer 64-bit capability top to bottom across our server product line-up.

Designed for systems based upon four or more processors, the platform includes the new 8500 chipset, which supports higher front side bus speeds, dual-channel DDR2 memories, PCI Express peripherals, and power management technologies that help lower utility costs. 8500 chipset also will support future dual-core Xeon processors along with Virtualization Technology, which will help users run multiple operating system environments simultaneously.

We announced that we are accelerating our Virtualization Technology Program by releasing a preliminary specification to developers and planning to include support in desktop processors and chipsets during 2005, 1 year earlier than previously expected.

Our 64-bit processor based platform, our 64-bit Xeon processor place based platform for volume servers has become the Company's fastest ramping server platform ever with approximately 2 million processor unit shipments in the first 8 months of availability.

During the quarter, we boosted the performance of our 64-bit Xeon processor for that platform by doubling the processors' cache memory with up to 2 MB.

In the Itanium family, Fujitsu launched its new Primequest line of enterprises featuring the Itanium 2 processor that combines datacenter class reliability and scalability with affordable standards-based technology.

We also announced plans to introduce a version of Intel Virtualization Technology for Itanium-based platforms later in the year.

Moving to our mobility group, we had record shipments of mobile microprocessors, wireless LAN components and application processors.

Notebook demand was quite strong across geographies, and mobile processors continued to grow a couple of points per quarter as a percentage of our clients' shipments.

In January, we launched our Sonoma platform for notebooks, which is based on new Pentium M processors, the 915 Express chipset, Intel PRO/Wireless LAN components and network connection software.

In the second quarter, we expect the Sonoma platform to almost double in volume and represent well over half of our mobile shipments.

Our flash business was down seasonally with lower units and slightly lower ASPs but was up 39% versus a year ago.

We remained the number one NOR memory supplier thanks to good density growth led by strong adoption of our multi-levels cell, StrataFlash chips in handsets.

During the quarter, we shipped our first 90 nm flash products and expanded our StrataFlash product line-up with new products for the embedded market segments, such as consumer electronics, industrial automation and wireline communications.

Unit shipments of application processors continued to grow rapidly, setting a record and nearly doubling from a year ago as designs from customers including -- Motorola, Palm, and Samsung -- have ramped into higher volumes.

Intel also introduced its first WiMAX component, designed for use in modems that connect homes and businesses to emerging WiMAX-based wireless broadband networks.

Formally code name's Rosedale, our new WiMAX silicon is being supported by 11 equipment manufacturers at launch and will be used by 14 carriers from around the world in commercial network trials.

In our Digital Home Group, we introduced a development platform for consumer electronics devices, such as IP-based digital set-top boxes and digital media recorders.

The platform is based on the 854 chipset Celeron end processors and software that helps speed the development of cost-effective and highly interoperable CE devices for the digital home.

Last week, Toshiba announced plans to introduce an advanced DVD player based upon this new platform.

We also acquired Oplus Technologies, a leading provider of video processing products and technologies for digital televisions and displays.

The acquisition provides Intel with technology that complements our growing portfolio of consumer electronics ingredients and digital home platforms.

In summary, I am pleased with our execution and performance this quarter.

We delivered solid financial results, introduced a number of important new products, and realigned the entire Company to develop platforms that better meet the needs of our customers and provide innovative and exciting new technologies for the marketplace.

We also remain on track with our industry-leading 65nm technology, which will begin production in the fourth quarter and will enable us to rapidly and cost effectively ramp dual-core processors in 2006.

Now, let me turn the meeting back over to Doug.

Thanks, Paul.

We'll now open the conference call for Q&A.

We will attempt to take questions from as many participants as possible.

To help in this process, we ask that you please limit yourselves to only one question and no more than one brief follow-up.

Thank you.

Operator?

(OPERATOR INSTRUCTIONS).

Adam Acree, Legg Mason.

This is Cody Acree but close enough.

Can you talk a little bit about where you may have seen some shortages in the quarter.

It sounds like as though availability may have improved towards the end of the period.

Was that the case?

And what do you expect for availability of some of those products into Q2?

Certainly, things did improve through the quarter.

So we ended it in a little bit better shape than we were in January and February.

I'd say the place you see the shortage or the tightness right now, most is in chipset, particularly the newer products and then a little bit in some of the mobile processors.

In general, we expect to meet all customer demand in the second quarter.

And Andy, can you talk a little bit about order linearity through the period?

How did things look towards the end of March?

And how were things as we pushed into April?

You know, we found nothing that was noteworthy throughout any of the first quarter.

But again, in a period when your tight, that's what you would expect.

We also though, as we got out and look at the world, we don't see alarming inventory builds.

It looks like what our customers are asking for is being packaged up and shipped on to their customers.

So right now, I think the demand we are seeing is real.

It is seasonal by the way if look at the first, second quarter.

First quarter was seasonal; second quarter we kind of expect to be seasonal.

So it is basically business as usual in a pretty good environment.

Glen Yeung, Smith Barney.

You know as a -- one of the big tech bellwethers in the world, we are seeing some weakness as we look around at other tech companies.

And just hearing your comments just now, I wonder if you can give us your sense on sort of the broader environment and the kind of confidence that gives you in your second quarter guidance?

Sure.

As we look through the world that we are going into, we see like I said, a pretty seasonal normal environment.

We do have a lot of sales that come from what I call the emerging markets.

Paul last week was in South America, saw strong demand there.

We see good demand -- the President of Intel India works for me.

We are seeing strong demand in that geography.

We see it in the South (ph) Asia.

But in general, if you look at the world-wide market, the world-wide economy, they are relatively healthy.

Are there going to be soft develop, we will see?

The economy is always an uncertain variable for us.

But as of today, world-wide demand seems pretty solid.

Great.

Thanks and just maybe as a follow-up, in talking about your CapEx being higher for the year, you suggest that it is a function of the supply/demand conditions you've seen in recent quarters.

Can you tell us how we should think about your utilization rates, as we look throughout the course of the year?

Do you expect that will remain tight for the preponderance of 2005?

Hard to say.

I guess the answer is probably.

I don't ever think I will get to the point where I'm extraordinarily comfortable.

I think it will remain an interesting year for us.

One thing I figured, there would be a lot of discussion on the capital spending.

So if I were to characterize it, I would say, probably three different factors driving it.

One is having been real tight in processors and chipsets for 6 to 8 months now, we came to the conclusion we were probably trying to run it a little too lean.

And we want to get a little bit of buffer in place.

Second, 65nm is looking good, and we are a little more confident investing there.

The other thing is, there's a lot of small projects.

But I do not want to leave anybody with the impression there is one big thing that is driving $500 million.

It's a variety of things -- a little bit of inventory buffer, a little bit of 65nm and a few other projects.

So when you say 65 is looking good, do you mean looking good in the yields are better or looking good in that you could be earlier with it?

Looking good in that I feel more comfortable getting four factories to ramp in it.

Adam Parker, Sanford Bernstein.

Paul, could you describe your competitive positioning in terms of price performance and architecture versus AMD and talk a little bit about how that has changed.

And do you think that GAAP between AMD and your chips has closed?

And when should I expect really Intel's desktop and server chips to be architecturally superior to AMD's again?

Well, that's a short question with a long answer, Adam.

But I'm not sure I have enough time to answer it.

Let me give you a quick summary.

First of all, I think the numbers speak for themselves in terms of the market dynamics between the two companies.

People pay a premium for Intel products, and we are outselling our competition by a large measure.

That is best indicated -- the market speaks the best.

Product by product, it's going to be different answers on different things.

In general, our approach is a platform-based approach.

And you have seen us take that very aggressively into the market in the form of Centrino mobile technology.

And that has given us a nice lift over the last couple of years, since we introduced the product, and is allowing us to I think outgrow those that aren't serving the mobile market nearly as well.

In servers, we have a very -- a quite refreshed product line.

You heard me talk about the 64-bit capability, now top-to-bottom, starting with Itanium to dual-core to MP Xeon and DP Xeon -- all with 64-bits now.

Again it is a platform-level architecture where the new chipsets are giving us platform-level performance that is keeping us very, very competitive in that market segment.

The desktop is really an issue of scale in scope.

And here our investments in emerging markets are what is paying off, where we were able to capitalize on growth in far-flung parts of the world that perhaps other competitors may not be able to cover nearly as well.

Okay.

So just in terms -- there's some notion that architecturally AMD had a gap, a lead a couple of quarters ago.

And your notion is that that gap is closed or is already closed?

Or can you make a price performance comparison between your chip in the desktop space?

I'm not going to go there, Adam.

Okay, cool.

All right, one follow-up then.

If AMD is going to exit the flash memory business, what do you suppose they are going to do there with the microprocessor tools once they are no longer really use of the micprocessors?

Or are they going to be competing in some of the areas like chipset and so forth more aggressively at that point or --?

I think their flash production was in a different set of factories than their microprocessor production.

Yes, it trended that way I guess a few years ago but --

You really need to ask Hector this question not me.

Okay, just thought I'd get your notion, given that you guys have been in the similar two businesses for the last few years.

There's not a lot of interoperability between the two.

The generation of technologies are the same in terms of N minus 1.

Increasingly, flash is becoming an N generation technology.

So it gives them a lot of excess capacity though a couple of years down the line.

Does that create capacity in areas that will be competing with Intel?

Or is that --?

Why don't you call 1800AMD?

One last housekeeping thing for Andy, I guess.

You had said that you had salary increases in the second quarter.

But I was trying to figure out with one less week of costs in this quarter sequentially, should we expect SG&A to be up on a sequential basis in absolute terms or --?

In theory, you would not.

But remember, I said there's two things.

There is the annual increases, and there's also increase in project spending.

So we should expect it to be slightly up in dollar terms then?

Yes.

David Wong, A.G. Edwards.

Your increased CapEx spend for 2005, could we reasonably assume that it actually doesn't have a very big effect on depreciation in 2006?

Depreciation for 2006 qualitatively might be pretty close to 2005 depreciation?

You actually sneaked a much bigger question inside your little question.

Certainly, the increase in spending in 2005 won't have a big depreciation effect on 2006.

What I would tend to do is take the whatever forecasts you had before looking at long term trend lines because depreciation changes very slowly.

And make very minor, if any, modifications there.

Okay, great.

And just one other clarification -- the increase CapEx, does that actually change the ramp plans for 65nm's in any way?

Does that actually bring it earlier into 2005 or is the original ramp point stay the same?

I suspect it has little effect on 2005.

It's more a matter of what we can do with it in 2006.

Jack Romaine, SG Cowen.

Could you guys help us understand some of the costs involved in the gross margin changes in terms of start-up costs -- what they were in Q1?

You mentioned a $100 million increase in Q2.

And could you also talk about under-utilization charges in your older 200 millimeter fabs that are producing flash.

Sure.

Let's do the flash one first; that is easy.

There are small under-utilization charges, but it really is a transition of the old process to the new process.

But as you ramp down on 135 or 108 nm and ramp into a 90 nm, you end up under-utilizing the old process.

So there's very small charges for under-utilization for that reason only.

Otherwise, the flash processors and the production is quite healthy.

If you start looking at crative (ph) costs, I realize it's always somewhat difficult.

So I'll explain the best I am able to give you what is going on.

Q2 will be $100 million higher for the 65nm start-up in Q1.

If you remember when we started here, I said the 2 quarters would be equal.

And in the update in March, we said the first quarter we found where we saved about a point's worth of margin, which is in $300 million neighborhood.

So savings in the first quarter going back up to original plans for the second quarter.

What we have said for the third and fourth quarter is, we expect start-up costs to return to a quote "normal" level.

A normal level is to 2 to 3 points of margin a quarter less than we see in the second quarter.

Okay.

And can we assume that under-utilization charges in Q1 were pretty close to 0, since you mentioned that the IA fabs were full?

There were not any IA -- the logic fabs.

There was that little bit on the flash for the transition from one factory, one process to the next.

Essentially assume the factories with that one exception for flash transition are loaded, and there are no under-utilization charges.

David Wu, Global Crown Capital.

Can you help me with two questions, Paul?

On the flash memory business, what is your concern about looking out that NaN and NOR seems to be comingled in the next-generation handset business.

You have to have these NaN-like characteristics thing and NOR to compete against the Samsung and Toshibas of the world?

And Andy on the 65nm, the yield seems to be good.

Should we expect lowered defect density and therefore better margins at any level of revenues in '06 than you did when you went from 130 to 90?

Let me try the flash one first, David.

In the part of the market that serves the upper end of the handsets, stacking in variety of memories is becoming increasingly common.

And we have been providing our customers with a variety of stacking options using our internal stacking technology to mix and match -- NOR, SRAM, EBS Rev. (ph), pseudo SRAMs, and now NaN.

And we have supply agreements in place to make that happen.

The configurations are almost customized by vendor by SKU.

And that requires that you have to have that capability to mix and match, as everybody in the industry will have to do I think over time.

And as for '06, our margin benefit from 55 nm, as you can imagine, I won't -- I will dodge that one.

What I will say is, we will show our '06 costs at the analyst meeting.

We are on track to continue to drive costs, as long as my factory stays full, we will be in pretty good shape.

Christopher Danley, JP Morgan.

Andy, I guess if you could just go through the sequential gross margin improvements -- how much is due to the lower unit cost versus mix versus the earlier call versus written down inventory?

Well, if you go from Q4 to Q1, I cited three things -- one is, lower unit costs; two is sales of privileged (ph) and improved inventory; and three is the fact that we had the one-time thank you bonus in the fourth quarter, which we don't have in the first quarter.

It is difficult; we don't give precise numbers.

But the first two are similar in stature; the third one is a bit smaller.

And then if we take out the extra week, was your sequential microprocessor unit growth pretty much normal seasonal?

I guess I am trying to reconcile the flat CPU units versus overall PCs sales that are down in the single digits.

Boy, again, it is hard to estimate what that week is actually worth.

I would say if you take out the week and are finding assumption, you would say, yes, it was pretty seasonal.

It is not too far off.

So it feels about normal then?

It feels normal to us.

Again, we're making an estimate as what the week is worth, but we are figuring the week is worth about half a week's worth of real shipments in revenue.

So as we try to triangulate it, yes, it feels pretty normal to us.

Michael Masdea, Credit Suisse First Boston.

I guess the first question is for Paul.

You talked a lot about pop (ph) from strategy.

Obviously, that's what you guys are doing there and restructuring also.

Centrino was clearly a consumer-driven phenomenon.

You did a great job of marketing.

Is it going to be a lot tougher to do this as you move into the enterprise and server side of your business?

Actually, Centrino was aimed for the first year and a half solely at enterprise.

We didn't begin our consumer ads until holiday season of '04.

And we did that on purpose because most notebooks were purchased at that point in time by businesses.

By the time we got to '04, we started seeing a spillover impact in the consumer markets, and we changed our advertising to cover both of them.

So I think that we are going to continue to focus on the conversion of business PCs from desktops to notebooks.

And we are architecting kind of end-to-end business solutions for the enterprise that I will talk about in more detail in our analyst meeting in a week or so.

You feel that the kind of opportunity for exploring chipset functionality is similar to what you see with the wireless for the notebook even though --?

Well, I think it's bigger.

I think the opportunity to use technology to lower costs and improve enterprise capability in the areas of manageability and security are much higher than anything we have addressed in the enterprise so far.

Great.

And then if you think about your business from kind of a return on investment perspective, you talked a lot recently in the last year or 2 about accelerating a lot of different things -- 64-bit dual-core.

It sounds like we are accelerating 65nm a little bit -- CapEx.

You are accelerating virtualization.

Are we changing the ROI model at all here?

Or is this just a competitive response?

How do you characterize what seems to be a lot of acceleration?

Well, I do not think we are changing the ROI model at all.

Each of the programs we do, we look at what we think the return is.

We don't make the investment unless we think we are going to get a very healthy return on these things.

I do not think it is changing.

Okay the acceleration that is coming -- you are looking at it; you are making a decision on ROI, and it makes sense -- what you're saying.

Absolutely.

I can almost say every decision, but I don't want to find a corner case I missed.

But every major decision is accompanied with ROI to make sure that we think we are getting a healthy return, and we'd have pretty high hurdle rates.

Tim Luke, Lehman Brothers.

Andy, I was wondering if you could give us some color on how you view inventory levels going forward and where you'd like to see that over the balance of the year in terms of days perhaps.

Boy, I do not know if I will do it in terms of days.

I'll do it in terms of dollars for you.

I still think my inventory levels are lower than I would like.

We are at -- we had $187 million in the first quarter.

I am still very tight on certain level processors.

I am still very tight on chipsets.

And some of the communication products, I'm hand-to-mouth on.

So I absolutely would like to see us get more in place.

One of the things we could go back and look at -- is it a mix problem?

Do I have too much of some and not enough of the other?

The answer is no.

One of the things we talked about was the shipping of previously reserved products.

You don't do that cavalierly.

As you're trying to meet your customers' demand and find solutions for them, you do go back and rework products.

You go back and look through the things that might meet somebody's needs and sell it to them.

So right now, I still need more.

To give you a precise number is difficult, but I'm not worried about building more inventory in the second quarter.

And certainly through the second quarter you would expect it to be higher then?

I would like it to be higher.

I can tell you that right now, my current forecast is a little bit but not much higher.

If I get some better yields out of the factories, then I'll start to feel more comfortable in the second quarter.

If I could ask a follow-up Andy, I was just wondering how you expect us to look at options expensing or whether you have a timeline on when you might start to factor that in.

Sure, we'll print options expensing into the -- we will follow the current rules, which say we will do it at the beginning of the year next year.

(multiple speakers) The first quarter?

Yes, first quarter.

We will continue to believe, as we have always believed, it's not an expense we put in the way we manage our business.

The efficacy has given us the ability to communicate to you both with and without, which we will do.

So we will actually file GAAP, and we'll let you know how we look at the businesses and how we think about it.

Joe Osha, Merrill Lynch.

Two questions -- first, looking at your dual processor desktop ramp here in the third quarter, can you give me a sense as to whether that's really going to be a low-volume relatively premium product?

Or by say the fourth quarter of the year, we could see parts volume for that part in the multiple millions of units?

Thanks, and I have a follow-up.

Well, you said third quarter, I think you meant second quarter in the first part of your (multiple speakers).

Yes, sorry.

Our goal is to ship to hundreds of thousands of units this quarter.

And I said in my commentary that our goal is to ship millions of units this year.

So you can extrapolate.

The reason that I ask is that certainly if you look at some of the third-party press, it's clear for multitasking and some high end users, there's a performance premium.

There seems to be some controversy as to for whether mainstream users running mainstream applications -- whether they'll be a performance premium at all.

Is your position net for the vast majority of users, there will be a performance uptick from single core?

Joe, it's going to be at an adoption curve.

There is some software out.

There is a lot of people that take advantage of the software today, typically in some of the gaming applications.

But we are talking about a PC market that is approaching 200 million units on a rounding rounder, right?

So millions of units is -- there is still a very small fraction of the overall market.

And the power users, the people who want that performance that is leading edge, easily fill up that number set.

Okay.

Great.

And then my second question is to do more with capital spending and the ROI issue that was brought up earlier.

Clearly, there has to be some kind of underlying demands assumption.

As you look through 2005 and 2006, is it necessary for revenue growth and business growth in this business to get back to where it was in '03 and '04 -- double-digit topline growth to realize an accessible return or hurdle rate return from this accelerated investment in equipment?

Sorry, I am not sure I understood the question.

Try me again.

Well, let me stand it on its head here.

You're accelerating capital spending in an environment where most estimates show revenue growths slower this year that it was in '03 and '04.

So I look at it, and I say, you were short in an environment last year where growth was higher.

Now you are accelerating capital spending in an environment where growth appears to be lower.

How do I know as an investor, that you're not just -- you are not top taking it?

We don't; we could be wrong.

We see first-quarter year-over-year growth of 17%.

The mid-point of our forecasted second quarter is 11%.

And we are expecting a seasonal half, and we have been short so far.

The good news is, even if we miscall it, the equivalent we're putting in place will be required for the next-generation process anyway.

So it's not like it goes to waste; it just means you bought early.

One of the things I hope to do at the analyst meeting is to give you a sense for how we do our ROIs on new process generations, so you can get a sense for the types of returns and the way we look at these things.

But it would take a pretty unusual set of events for us to buy this extra capital and have it end up not getting a return.

Krishna Shankar, JMP Securities.

For the dual-core processor ramp, can you talk about the types of applications in the corporate market that would need dual-core processors?

And secondly, when do you anticipate to introducing dual-core mobile PC processors?

I think that most of the product, the desktop dual-core products, sold this year are going to be into the consumer power user and some workstation applications.

Certainly at workstations, the software is already threaded; the operating systems are already threaded.

The applications are there.

So that is an easy fit in.

But even in some of the basic consumer applications by running background tasks, you can get 40, 50, 60% performance increase, depending on the application environment.

So we will start in consumer, some of the specialized workstation environments, and we will drive it with the software over time into the overall marketplace.

We are -- the first mobile product is a product that is -- that dual-core is a product that is code named Yonna (ph).

We have demonstrated that product at our developer forum.

And now -- and that's on track for revenue shipments this year.

John Lau, Jefferies & Company.

In moving over to the dual-core and 65nm's, you had an opportunity to reduce your clock speed by the increased performance, and the heat came down.

That was a similar wrote on the Pentium M. Is the process limit that you are doing right now, that 4 GHz market, is that a barrier right now that you are trying to utilize architecture to really improve the performance?

And I have a follow-up, thank you.

In general, we're trying to balance the overall what I call "price performance per watt."

So we look at the overall ratio of how much performance we can get out in a given thermal envelope that certainly in notebooks, leads you to taking a lower clock speed implementation using more architectural feature to deliver performance like we did with Pentium M and now with the dual-core stop.

As we move to dual-core in the desktop though, particularly in the Extreme editions, you will see us continue to drive the clock speed within a more aggressive power envelope that those machines can handle.

So I don't think the 4 GHz is necessarily a fundamental level.

Just the envelope you are trying to work in?

The thermal envelope is the key.

I do believe the thermal envelope will come down over time even in desktops.

And the final question is -- focusing on your 65nm and the fact that you are ramping faster, what is your target percentage of production on 65 nm for 2006 on microprocessors and also what is that percentage for flash on 90 nm in 2006?

Thank you.

Let me just correct you on something you said upfront.

We are not ramping any faster than we said or planned, right?

We are still ramping exactly on the schedule we have been talking about for some time now, for quarters.

In terms of the mix, 65 versus 90 a microprocessors -- I think we will save that and do that update in the analyst meeting -- and probably do a similar one for flash on the 90 nm crossover.

Ben Lynch, Deutsche Bank.

Andy, first a question for you.

I had noted that on the mid-quarter update, you said something like, maybe we will be able to grow inventories expecting to slant up a little bit now versus that inventories are up 7%, and you had better sales, which would presumably made it difficult to grow inventories -- it seems like this improvement relates to manufacturing efficiencies?

Or is it something else?

And assuming there are manufacturing efficiencies that also benefit gross margin, could you go into what those are?

And I have a follow-up please.

There are a couple things.

First of all, if you recall, I said some part qualified earlier, which is a good thing.

So think of -- is both -- there's a microprocessor and a chipset product that we expected not to qualify until next quarter.

Instead, they qualified in the first quarter.

And what that means is, you can value the production and put on the balance sheet instead of having to expense it in the quarter.

So our goal is to build some, which we succeeded on, and then we placed some more value onto the balance sheet because of it.

It does help margins; there is no question that it is one of the things that improve margins.

And it is for that.

We get some benefit out of factory efficiencies.

We are getting good yields; we are running full factories.

We're getting a little bit more benefits to the bottom line.

We get lower cost units because of that.

So that's all happening a little better than expected.

But there isn't any one thing that really stands out other than the qualification of the product -- getting the products ready for a revenue shipment early.

Okay.

And then you clearly did well in Q1 on gross margins, led you to raise at the full year gross margin target.

When you look at the other sort of swing factors that you have that excludes just revenue upside --- are there still other swing factors?

What is sort of the major things that could lead to further margin upside?

And where is still the perceived risks that again exing out how revenues may evolve through the year?

Certainly, revenues have a lot to do with this, right?

It is hard for me to x out --

Yes, I know.

Probably the biggest swing on our margin up-and-down or costs up-and-down is how full we keep the factories.

But if you assume the factories stay relatively full, the things then that can lead to outsider -- the same things we've seen.

You see a stronger mix to fixed processes.

You see a stronger mix of Pentium.

You see a stronger mix to mobile.

You see a stronger mix to server.

Those things will all tend to help the bottom line.

If you get a little bit quicker ramp on a couple of other products that is coming out, or the products and the new processors, that will tend to help your margin overall for the year.

The things you worry about, you worry about field crashes.

You worry about all the stuff that you don't know about that can go wrong.

That right now, if you ask my opinion, I would say, if you give me a strong revenue year, the margin will take care of itself, and I would be pretty comfortable with that.

Okay, Operator, we will take two more questions, please.

Hans Mosesmann, Moors & Cabot.

65 nm -- what are you expecting in terms of thermal envelopes with that process relative to 90 nm?

I recall that you had some issues with that one at the beginning of its ramp.

What is the expectation for 65?

The initial products will fit inside both the mobile and existing desktop power envelopes.

And then we will try to bring out derivative versions that take this down over time, as we crank up architectural features and processes, so forth.

But nothing -- we're certainly not planning to take it up.

And just a follow-up --

Remember, these are all dual-core products.

So that is a fundamental shift there.

That is helpful.

And then a follow-up in terms of 65nm, you said you weren't ramping faster than expected.

Is that true?

We're not ramping faster than expected this year.

We are prepared to be able to take better advantage next year.

Mark Edelstone, Morgan Stanley.

Good afternoon guys, and nice job on the margins.

Andy Bryant, just to clarify a couple of things you said on the margins in the core;

I just wanted to make sure I was about right on it.

The sale of previously reserved inventory, it sounds like that was maybe 120 to 150 basis impact in the quarter, is that about right?

No, that is too high.

Okay, and then the start-up costs, they go down in Q3.

Should we be expecting that from Q2 to Q3 -- you're getting a 200 to 300 basis point improvement all other things considered equal?

Yes, all other things considered equal, we should see 2 to 3 points going from Q2 to Q3 for start-up costs alone.

Great.

And then just lastly, maybe for Paul, I think as you guys were going through Q1, things were even tighter from a supply-demand perspective.

My understanding is that you came off of allocation, or at least things eased up a little bit on some of the product portfolio.

And as you were coming off of some of the tighter supply-demand environment, can you just give us a sense as to what you saw with orders and just overall backlog volatility?

Really no significant backlog volatility or change in order pattern, Mark.

And the other indicator I would put in here would be the gray market.

There's not things -- they are selling at a significant premium or discount right now.

The fair market pricing tends to hang right around our list pricing.

Okay.

We would like to thank everyone for listening to today's call.

A recorded playback of this call will be available at approximately 5 PM Pacific Time tonight.

Those interested should dial 1-800-642-1687 and reference passcode 5281775.

Thank you.

This concludes today's Intel Corporation first-quarter 2005 conference call.

You may now all disconnect.