英特爾 (INTC) 2004 Q2 法說會逐字稿

Good afternoon, my name is Amy and I will be your conference facilitator today.

At this time, I would like to welcome everyone to the Intel second-quarter 2004 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.

If you would like to ask a question during this time, simply press star, then the number 1 on your telephone keypad.

If you would like to withdraw your question, press star, then the number 2.

Thank you.

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

Okay.

Thank you and welcome to the Intel second-quarter earnings conference call.

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

I'd like to remind everyone that the earnings release in this call are available on our IR website at intc.com.

For those of you who did not see the earnings release, revenue in the second quarter was $8.05 billion and earnings per share were 27 cents.

The second-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 July 12, 2004.

Lastly, if during this call we use any non-GAAP financial measure as defined by the SEC in Reg G you'll find on our website, intc.com, the required reconciliation to the most directly comparable GAAP financial measures.

And I now want to introduce Andy Bryant, who will discuss second-quarter earnings results.

Andy.

Thanks, Doug.

In the second quarter we continued to achieve strong year-to-year growth in revenue and profits.

Revenue grew by 18%, gross margin was higher by nearly 9 points, and operating income was up 87%.

Overall sequential results were flat at the high end of what we have seen in the past second quarters as compared to first quarters.

The major positive in the quarter was Flash memory, with strong growth in revenue and profit.

We look forward to a good second half, although as economic recovery -- recovery proceeds, the year-to-year growth rates of the last year will inevitably subside somewhat.

For the full year, we have lowered our forecast for gross margin which we expect to approach historic annual highs.

Revenue of $8.05 billion was approximately flat with the first quarter, a little below the midpoint of our June update and a little better than the pattern typical for this period.

Revenue for the first half is up 19% from the first half of 2003, the best year-to-year growth for the first half we have seen since 1997.

Unit shipments of Intel Architecture and microprocessors were lower than the first quarter, while unit shipments of chipsets, motherboards, Flash memory, and connectivity products were higher.

Revenue in the Intel Architecture business was $6.8 billion, down nearly 4% from the first quarter and up 16% from a year ago.

Within this group, microprocessor revenue of $5.8 billion declined a seasonal 4% from the first quarter and grew a substantial 19% from the second quarter of 2003.

Most of the remaining revenue in the Intel Architecture business comes from chipsets and motherboards, and amounts to $1 billion, approximately flat with the first quarter of this year and the second quarter a year ago.

Revenue in the Intel Communications Group was $1.3 billion, 19% higher than the first quarter, and 33% higher than a year ago.

The greatest portion of this growth came from Flash memory, whose sequential and year-to-year growth was the best it has been in years.

Revenue from wireless connectivity and the communications infrastructure products was up nicely.

Gross margin dollars of approximately 4.8 billion were 2% lower sequentially and 38% higher year to year.

On a percentage basis, gross margin was 59.4%, just below our June forecast of 60 to 61% and just below the first-quarter level of 60.2%.

There are a number of items to highlight in reconciling the change in gross margin from the first quarter to the second.

The first quarter included a charge of $162 million related to a legal settlement which lowered reported gross margin percentage for that period by 2 points.

The second quarter included a warranty reserve, we did not anticipate in the June update, related to a manufacturing excursion affecting certain components of chipsets.

This amounted to about 1.2 point of gross margin percentage for that period.

The net impact of these two charges from the first quarter to the second quarter was an improvement of about 1.5 points .

More than offsetting this were other factors lowering gross margin by a couple of points.

The primary factors were a mix shift in revenue to lower margin products like Flash memory and motherboards and a very slight drop in pricing of some microprocessors.

Gross margin percentage was up nearly 9 points from the second quarter of 2003.

Operating income of $2.4 billion was down slightly from the first quarter and up 87% from the second quarter a year ago.

In the Intel Architecture business, operating income of $2.8 billion was 41% of its revenue.

This is a sequential decline of 7% and a year-to-year growth of 52%.

The communications businesses cut their losses to $126 million, nearly half the loss of the first quarter and half the loss of the second quarter of 2003.

Higher revenue from Flash memory and lower inventory reserves were the key drivers of this improvement.

Spending on R&D, marketing, and G&A was $2.4 billion, consistent with our forecast and approximately flat with the first quarter.

Although up 12% on a year-to-year basis spending has declined as a percent of sales, the number of employees at the end of June was 81.7 thousand compared to 80.5 thousand in March, and 78.7 thousand in June of 2003.

The total in the second quarter for interest income, other income, and gains and losses on equity investments was $39 million, lower than the $60 million we anticipated in June due to higher impairment charges.

In this category, interest and other income was $47 million.

The effective tax rate for the quarter was 27.4%.

Second-quarter results included a previously disclosed reversal of $62 million of accrued taxes, as well as an adjustment made to reflect an increase in the estimated tax benefit for export sales.

These items increased second-quarter earnings per share by 1.7 cents.

Diluted earnings per share, which include potential dilution attributable to employee stock options, was 27 cents.

Basic earnings per share, which does not include potential dilution, was also 27 cents.

Average shares for calculating diluted earnings per share were 6.6 billion, approximately flat with the first quarter.

During the quarter, we repurchased 56 million shares at a cost of $1.5 billion.

Basic shares outstanding at the end of the quarter were 6.4 billion, down 1% from a year ago, and down from the peak in the second quarter of 1998 by 5%.

On the balance sheet, accounts receivable were down $191 million or 6% from the first quarter, reflecting a quarter that was less back-end loaded than the prior quarter.

Day sales outstanding were flat at 36.

Inventories increased by $427 million, by 15%.

A faster rate than we anticipated in April.

The largest increase is in work in process and finished goods.

Almost half of the net addition to inventory is from microprocessors.

Driven by our new 90-nanometer microprocessor made on a 300-millimeter process often referred to by the code name of Prescott.

The yields on this product have been better than we forecast and our inventories have been higher.

The other half of the increase is split between Flash memory and chipsets.

Higher levels of these products reflect higher expectations for demand.

Starting this quarter, we will work to lower the inventory levels of microprocessors by slowing the planned growth rate and wafer start of these products.

Although the higher than anticipated yields on the new processors are encouraging, the resulting actions to reduce inventory will lower expected profit margin in the months ahead, which I will discuss more when I review the outlook.

Cash, short-term investments, and fixed income trading assets ended the quarter at $16.7 billion, an increase from the first quarter of about $1.2 billion.

After the stock repurchases of $1.5 billion, capital spending of 1 billion and dividend payments of 258 million.

As we turn now to the outlook for the third quarter, please keep in mind that the forecast data does not include the effect of any new acquisitions or divestitures that may be completed after July 12.

I'll use the midpoint of forecast ranges in making comparisons to prior periods.

We expect revenue in the third quarter to be between 8.6 and $9.2 billion.

The midpoint of this range would be sequential growth of 11%, which is at the high end of patterns typical for the third quarter.

In the five years prior to this, growth in the third quarter has ranged from 3 to 15%, and averaged 7%.

Compared to the third quarter of 2003, the midpoint of the forecasted revenue range will represent growth of 14%.

Our plans assume that revenue from microprocessors will grow at a slightly faster rate than that typical for the third quarter.

We also expect revenue from Flash to grow at a faster pace than what we would normally expect as we plan to increase market segment share.

Gross margin percentage in the third quarter should be 60% plus or minus a couple of points.

This compares to 59.4% in the second quarter.

If we continue to succeed in gaining share in Flash, chipsets, and motherboards and these products grow at a faster rate than microprocessors then our overall revenue mix will have the greater proportion of lower-margin products.

This in turn will tend to lower overall company gross margin percentage.

At the same time, while margins on microprocessors are forecast to expand during the quarter, they will not increase as much as I would have expected based on the revenue increase and our current outlook.

As I mentioned before, we have slowed the rate of growth of wafer starts to reduce inventories, which will slow the rate of reduction in unit costs.

At the same time, we expect modest downward pressure on prices.

All of this translates into good growth and revenue and gross-margin dollars, but flat gross-margin percentage.

The midpoint of 60% is approximately 2 points higher than the gross margin in the third quarter of 2003.

Spending.

R&D plus MG&A should be approximately $2.5 billion, a slight increase in second-quarter spending of $2.4 billion.

Depreciation should be between $1.1 and $1.2 billion.

Amortization of acquisition-related intangibles and costs should be approximately $40 million.

Our estimate for gains and losses from equity investments and interest and other income is a net gain of $50 million.

The tax rate for the third quarter is now expected to be approximately 31% revised from the prior outlook of 32%.

Looking ahead to the rest of the year, we are planning for seasonal growth in microprocessors, and exceptional growth in other product lines.

Although our expectation for profitability remains high, we have tempered down to reflect current trends in unit costs and pricing for microprocessors and higher growth for Flash and other products.

We now expect gross margin percentage for 2004 to be approximately 50% plus or minus a couple of points.

Our previous forecast is in the midpoint of 62%.

Rate of reduction in unit costs will be slower as we start fewer wafers than we had planned but the impact on both costs and inventories will be greater in the fourth quarter.

The inventory reduction in the third quarter will most likely be minimal, but with the seasonal second half we should see inventory declines in the microprocessor product line of hundreds of millions of dollars by the end of the year.

The 60% midpoint is over 3 points higher than 2003 gross margins of 56.7%.

Our forecast for capital spending remains $3.8 billion, plus or minus $200 million.

Our budget for R&D for 2004 is also unchanged at $4.8 billion.

Depreciation for this year is forecast to be approximately $4.6 billion.

Amortization of acquisition-related intangibles is expected to be 175 million for 2004.

In conclusion, the solid microprocessor business is tracking at a healthy growth rate, while the communications businesses, led by Flash memory, are showing signs of renewed vitality.

Our new manufacturing processes are healthy.

For the full year, we are on track to deliver double digit growth in revenue with high levels of profit and cash generation.

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

Thanks, Andy.

The second quarter saw seasonal results for our IA business along with good growth in our communications group, led by a rebound in Flash and continued growth in wireless LAN.

It was also a significant quarter for us with respect to new product announcements as we introduced our first 90-nanometer products for Mobile and the Enterprise along with a major platform upgrade for the desktop.

On a geographic basis, our strongest region continues to be Asia-Pacific, with growth of 11% sequentially and 32% year-over-year.

China set a new revenue record with good demand in both consumer and corporate market segments.

We also had healthy growth in South Asia, including India.

Handset strength and Flash market segment share gains contributed to APACS growth this quarter.

In the Americas, revenue declined 10% sequentially, and was flat versus a year ago.

As the PC supply chain moves more production to Asia-Pacific, some of our sales move as well.

Understating consumption growth trends in the Americas region.

Our data suggests continued healthy year-over-year growth in both U.S. retail and corporate, especially for servers.

In EMEA, revenue declined 14%, in line with normal seasonality and grew 17% year-over-year.

We saw strength from our large OEM customers and weaker than expected demand in the channel.

We also saw good Flash growth at our customers in Europe.

In Japan, revenue grew 7% sequentially and 15% year-on-year thanks to good demand for our communications products such as Flash and wireless LAN.

In our Intel Architecture business, we saw increased exports of notebooks but a seasonal decrease in local PC consumption.

Moving to our product groups, Intel Architecture revenue was down seasonally but grew 16% versus a year ago.

Microprocessors units were down slightly and ASPs were down a bit.

We believe we held market segment share in the second quarter and significantly strengthened our product lineup with several major product launches.

Our desktop business experienced seasonal weakness in the second quarter but had double-digit year-over-year growth.

We continued to ramp Prescott and reached crossover with Northwood in the performance segment in late June.

Total shipments of Prescott in the quarter were lower than we projected a quarter ago, as our customer conversion ran a few weeks behind our goals in the early part of the quarter, but accelerated in the second half of Q2 to very healthy levels.

We expect to more than double shipments of Prescott in Q3 as platform costs continue to drop, availability of high-frequency parts increases, and the market benefits from the compelling new platform features of our new Grantsdale chipset.

Grantsdale, which was introduced in mid-June, is shipping in high volume and delivers some of the most significant PC platform enhancements in a decade.

Including support for high-definition video, 7.1 surround sound audio, dual displays, and redundant hard disc drives.

Our mobile business delivered higher units led by solid demand for our Centrino mobile technology.

Which continues to grow as a percentage of our mobile units.

The highlight of the quarter in Mobile was the May launch of our 90-nanometer Mobile processor code-named Dosan (ph).

We expect this product to ramp quickly, crossing over Banias during Q3, only one quarter after launch.

In the enterprise, we had higher units led by ongoing demand for our dual processor and multiprocessor server products.

Revenues were up more than 40% year-over-year for the second quarter in a row.

In June we launched Nocona our first enterprise processor built on 90-nanometers.

Nocona based workstation platforms are on the market today, and server introductions are expected in a few weeks.

Our Nocona platforms bring higher levels of performance and new capabilities to the enterprise.

Including PCI Express IO and graphics, DDR2 memory and 64-bit memory addressability.

We expect a very rapid transition to Nocona in the dual processor segment over the course of the second half of this year.

In June it was announced that more than half of the 500 fastest super computers in the world are now based on Itanium or Xeon processors, reflecting the trend in super computing away from proprietary risk architectures.

The Itanium 2 processor in particular has seen strong adoption in super computers this last year, more than tripling the 61 systems in the current top 500 report.

In chipsets, we had growth in both units and billings, as we regained market segment share from the first-quarter levels.

Our plan in the second half is to continue this momentum as we ramp Grantsdale in the high end and mainstream segments and move our previous generation products into the value segments, providing us with a strong product lineup from top to bottom.

Moving to our communications business, revenue for ICG was up 19% sequentially, and 33% year-over-year led by strong Flash demand.

Flash revenue grew 41% sequentially to $587 million as we successfully reengaged with the broad market and saw continued strength in our high-density products for handsets.

We believe we gained several points of share in both Nor and the overall Flash market segments.

We also began sampling 90-nanometer parts to customers, with volume productions scheduled to begin in Q3 and revenue shipments by year-end.

Total Ethernet -- total Ethernet connectivity units set a record thanks in part to double-digit sequential growth in our wireless LAN products.

Our 80211 BG product, which we introduced in the first quarter, is now our highest volume Wi-Fi product.

In applications processing, customers including Dell Computer and PalmOne introduced new PDAs based upon the recently announced Intel PXA270 family of processors.

Formerly code-named Overti.

In summary, for the quarter, we maintained strong year-over-year growth and delivered a stronger product road map in Desktop, Mobile and Enterprise that will serve us well as we enter the growth half of the year.

Our 90-nanometer process is ramping to become our mainstream CPU process for the second half of '04 and we plan to use our capacity to deliver a continued performance in volume leadership in microprocessors, and to grow our presence in chipsets, flash and other communications components.

With that I'll turn it back over to Doug to begin the Q&A portion of today's call.

Okay, thanks, Paul.

We will now open the 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.

As a reminder, ladies and gentlemen, if you would like to ask a question, please press star, then the number one on your telephone keypad.

Once again, ladies and gentlemen, that's star, then the number one.

Your first question comes from John Lau with Bank of America.

Thank you.

I was wondering if you can give us some more color on the -- on the signal that you are getting that allow you to guide above normal seasonality for Q3 and what are the different areas that you are seeing that strengthen?

Thank you.

Sure, but we need to be careful somewhat.

If you look at the total company Q3, the growth rate that's in the midpoint of that is higher than you would normally seasonally see, I agree.

If you look at the pieces, what you see inside microprocessors are something close to the middle maybe a little above the middle in terms of the growth rate, kind of a normal seasonal third quarter.

The reason it's growing more sharply is you saw the growth -- strong growth in Flash, Paul talked about in the second quarter, 40% quarter to quarter, we think we'll continue to take share in Flash.

We think we'll take share in chipsets, we think we'll take share in motherboards.

So we're looking at our third quarter as a seasonal, maybe a little better but not much, kind of typical Q3 with an opportunity to take share across a broad range of our products.

Now, what are we seeing that gives us comfort about that kind of seasonal Q3?

Well, Q2 for example, it behaved in the Intel Architecture space about as perfectly seasonal as it could have.

If you look at -- well, I talk to our customers, we get input from them.

We look at the worldwide economy, China set a record for us in the quarter.

Everything out there is behaving right now normal to -- normal to slightly positive, take that as your baseline, add in the market segment share and the other products and that's how you get to a stronger than normal quarter.

Okay.

So as we move from the normal seasonal quarter in Q2 and we see out there can you now project what you see in Q3?

Nothing deters you from any weaknesses or any other areas that would be lower than seasonal at this point?

We haven't seen anything yet, talking to customers we haven't seen that, looking at worldwide economies we haven't seen it.

The thing we have seen of course is the software companies during the preannouncement which gives you a little cause for concern but inside our business we're seeing no evidence yet of a problem.

Great.

Thank you.

Your next question comes from Adam Parker with Sanford Bernstein.

Yeah hi.

Exactly when did you guys decide you have to slow the planned growth rate of processor wafer starts?

You know, it seemed like as recently as your analyst meeting in May in New York, it wasn't something that was on the table, so did this change during the last few weeks of the quarter?

Even later I'd say it changed at the very end of the quarter.

So when we -- when I talked to you in June, we had, the foot on the accelerator.

I'd say we got towards the last part of the calendar month, June and that's when we do our whole year forecasting for the rest of the year.

We looked at it and said it's time to make a change and, you know, Paul placed the order at that time and everybody then started changing the build plan.

Retrospectively, do you think this could have been done earlier in the game, or -- because it seemed like in the Q1 report when inventory built -- you know, 11% or so sequentially and all the entire technology stocks got killed on inventory concerned, you know, people got -- asked you a lot of questions on the call, hey, is this going to impede the back half and it felt like you guys thought there was a low risk of that.

Is -- is it something you -- you know, you could be more proactive about in the future?

Or how do you -- how do you make that decision?

With hindsight I can certainly look at it and say I wish I'd figured it out a month or so earlier.

What we always are toying with, Adam, is in our particular business, the cost of missed sockets or missed sales is enormous.

And so, yeah, we took more risk and ran the factories longer, in reality as I look back now I wish I'd pulled them back sooner.

So this is basically what you guys used to call an underabsorption charge to cogs, you're really taking in the back half of the year, is that fair?

I mean the last time you did this was what?

Q2 2002.

Yeah.

It's actually we're not pulling it down after being under absorption but it does mean the amount of depreciation absorbed by each part we build will be higher so I don't get to reduce my cost as much as I had hoped for.

Can you help at all with magnitude on a dollar size, is this like a $200 million penalty to cogs in Q3 or what's the order of magnitude based on the idling of the factories?

I -- I can't give it to you that way but let me try something else.

When we look at the full-year margin, we were at 62 plus or minus a few now I'm going to be at 60 plus or minus a couple.

So the midpoint moves about two points.

Right.

If you're looking at that movement of midpoint, one of -- half of that comes from, the increase in revenue in the lower margin products.

So I get more absolute gross-margin dollars, I get a lower percentage.

The other half comes from a combination of not getting my unit costs down as low as I would like and a little bit of price pressure.

Let me go now, I want to correct one word Adam that you used the word idling.

It's not -- ramping up more slowly, is that --

Right.

We're not idling anything, we're just not ramping it quite as fast as we first thought.

On the other hand, the effect is the same.

We still expect to exit the year with the vast majority of our microprocessor line top to bottom on 90-nanometers.

Great.

So doesn't this have contagion beyond this year?

I mean, in other words, your plan to build capacity seems like it -- you know, it has to alter -- be altered beyond just the back half of this year.

It is something you do have to watch closely, but, remember, as Paul said, by the end of this year we're at 90%, on the 30 -- 300-millimeter 90-nanometer set.

So doubling in Q3, substantial growth in Q4.

So I need to use those new factories out in that period of time.

One of the reasons you see us getting more aggressive in chipsets is because it frees up some of the 8-inch capacity that was being used to produce the -- the Northwood.

So what we have to do is look at that trailing capacity, put wireless communication chips in for Flash and put, chipsets in.

So you're going to see us be more aggressive in those spaces because we think we have the capacity to take advantage of it now.

If demand is lower on a year-over-year basis next year than it is this year and you're still increasing at the same rate, doesn't this risk of kind of a underabsorption or -- or, you know, stalling the ramp up persist for longer than just the back half of this year?

It -- I would say depending on how much ahead of time you figure it out, it's the -- a yes or no question.

If you knew that in 2005 demand was going to be soft, what you would tend to do then is drive more logic product into the 12-inch environment so you still use your best technology, and you would end up with excess older 8-inch factories.

Okay.

So --

That --

At this point your judgment is what?

Well, again, you're talking about next year, which I'm not supposed to be making a forecast yet.

Our judgment, quite frankly, is our capacity is just fine, we have about a four- to six-month lull, if it turns out demand is softer next year, we will drive more products into the 12-inch factory and we'll have an extra 8-inch factory.

It just seems like, you know, you guys went from saying, you know, a couple months ago that there was a -- you know, less than 50/50 chance I think was your phrase or, you know, pretty small chance that you would have to, you know, consider this to -- to doing it, you know, and it happened -- you know, as you said all of a sudden or, you know, at the end of the quarter and I just -- I'm trying to figure out how we -- you know, how we reconcile this in terms of the tremendous amount of capacity coming on-line and sort of a general sense that the U.S.

PC demand's going to be lower next year versus this year.

I'm not sure what question you're asking.

I'm just saying, you know, --

Are you asking do we -- were you seeing it sooner and we were surprised?

Yes.

Okay.

So it just -- it came as a total surprise and you basically at the end of the quarter had to say okay this is the time to do it.

We wouldn't say total surprise, you were watching it, you're seeing the yields get better.

As they continue to get better we got more inventory, we said we need to -- we need to not idle, we need to slow the ramp down a little bit.

At the same time we needed to make sure we can use those factories into the future, what -- the most important thing is use your 90-nanometer, use your 300-millimeter and then try to find a way to fill the older factories.

All right.

Thank you for your time, thanks.

Your next question comes from Glen Young with Smith Barney.

When you look out -- or look back into the quarter we just finished, not -- you don't have to be too specific on linearity, but was there anything from a demand perspective that changed towards the end of the quarter that may have resulted in some of the inventory concerns or is it purely a function of the ramp?

Not over the quarter, relative to the quarter per se, but in the second quarter we always -- you know, as we approach the end of the second quarter you always get a much better visit -- view into the second half of the year.

April's a very difficult time to judge Q3 and Q4.

June is not perfect but it's a lot better than April.

So -- so our view of the second half is really what allowed us to integrate, the -- the yields out of the -- the 200-millimeter factories, the overall demand and the inventory situation to make the decision that Andy just discussed.

Is that to suggest, then, that your view of the second half was actually higher than it is now given that you are reducing the wafers?

Well, as I said, there's a number of factors here, one is demand, one is yields and one is inventory.

As they all came together, we felt this was the best mix.

Okay.

And then just one other thought there on the mix issue which is if you look out into -- you know, into the mix from a desktop versus notebook perspective, is there anything that's -- that's unexpected going on there or is it the mix as you pretty much expect to see?

No real shift there.

Okay.

Thanks.

Your next question comes from Cody Acree with Legg Mason.

Thanks.

Maybe to move to a different topic.

Can you talk about the Grantsdale manufacturing issue?

Can you talk maybe about the adoption targets now for the end of the year and maybe how those have changed over the last few weeks?

The manufacturing issue has -- was a -- an excursion - in one of our factories building that chipset.

It was being produced in more than one factory.

The parts in that one factory -- one of the complements in that one factory had a manufacturing issue that caused us to pull those products back and replace them.

We have now done all of that replacement and we are currently meeting all demand on Grantsdale.

So I don't really see any impact second half relative to the sort of one-week privation (ph) we had with that recall.

No material delay in any of your customers' ramps or any stalls --

No.

We had sufficient --

In the market to solidify the Grantsdale ramp?

No.

We had sufficient product in the line and behind it and in inventory to be able to do the swaps pretty quickly.

Are you still targeting 50% by the end of the year?

That would be -- that's where -- is our goal.

Great.

All right.

Thanks, guys.

Your next question comes from Chris Danely with JP Morgan.

Thanks, guys.

So I'm just trying to sort of read in here as to what happened in June.

You know, the number -- or the revenue was a little below the midpoint and it sounds like, Paul, you gave the lower utilization rates order late in June.

Did you guys see any kind of fall-off in the last couple weeks of June like we've seen with some of the other tech companies?

Not substantially different.

You know, if you look at the second quarter, June was necessary to get to 80, 50 and it came in fairly well.

Did I expect it to be 81?

Of course.

So it means I got a little bit less than I expected but nothing significant and no real change in shipment patterns.

Okay.

So maybe a couple of push-outs but no big deal?

Yes.

And then on the -- the lower -- the lower unit ASPs for processors in Q3.

What's the primary reason for that?

Is that just to help in clearing out inventory?

No, not really.

You know, we're not changing our price stack based upon the work in process inventory that we have.

It's really the -- where we see the markets and the overall mix of products.

Right now, as I indicated, our merging markets are pretty hot, they tend to be more desktop centric.

And that -- that puts a little bit more pressure on the average selling price and the real issue is, you know, does the servers continue to grow and notebooks continue to grow to offset some of that.

So -- so it sounds like if -- if demand is higher on the lower-margin products, particularly I guess on the Flash chipset motherboard, Paul, why -- why go after the Flash share and the chipset share and the motherboard share versus just pulling back there and focusing more on processors?

Because we're in this business to grow our overall profit dollars, not the percent.

Got it.

Thanks.

Your next question comes from Michael Masdea with CSFB.

Yeah.

I guess -- sorry just one piece here that the slower than expected cost reductions, your 300-millimeter, 90-nanometer seems to be ahead of plan and your yield seem to be better than expected.

It seems like that would suggest your MPU costs are actually better than expected, can you help us just rationalize that?

Repeat that one again for me.

Yeah.

You said that your cost reductions for MPUs were slower than expected.

Yes.

And, you know, earlier -- last quarter you told us 300-millimeter, 90-nanometer was ahead of plan and then this quarter you told us yields were a little bit better than you thought.

It seems like those two would lead to better than expected cost reductions?

Yes.

So what happens as we start fewer wafers throughout the rest of the year than we had expected it means each product now gets a bigger piece of depreciation and a bigger piece of labor and overhead than it otherwise would have.

So if I could have continued to run the factories that pull out and ramp them as fast as possible I would have made my costs goals without any trouble at all.

So costs are still coming down just not as much as we could have gotten them down if we had kept the factories ramping full out.

So it's not a specific cost issue to the design or anything else, it's just an inventory issue.

Oh, no.

It's just a matter of --

Dollars and units.

Right.

I guess it's as simple that is.

Okay.

That's fair.

The second question is on the Flash side, is it a situation where you're gaining share because you see lead times pretty tight across the board or do you feel like you've really actively re-engaged with the customers now?

The principal upside of the last quarter was our re-engagement of the broad market through the distribution channel.

We've had -- I talked about some of the success in the cellular handset market, but, the real significant numbers in terms of growth were -- were the channel re-engagements.

And where do you stand on the OEM front right now?

Going.

Okay.

The final question just on the customer side, you said patterns haven't changed a whole lot, is there any kind of change at all coming from your customers or were you expecting anything different than what you're seeing?

Boy, nothing -- nothing out of the ordinary.

And -- and it -- what gives us comfort I think is that it's very consistent on a worldwide basis.

The channel, the OEM, geography by geography, we're all getting a very similar message.

I guess just the last question on the ASP side since you talked about the mix, what do you have to combat that.

Last year you had Centrino which helped a lot.

You had HT kick in a little bit.

Is there anything out there that seems to be able to combat what you're seeing on the mix side from an ASP side?

Well, I mean I think the biggest issue is -- is the mix of desktop versus notebooks and servers and that really is a function of which countries and which markets jump fastest over the second half of the year.

That's why we gave you that -- you know, I think some conservative guidance on second half ASPs.

Great.

Thanks a lot.

Your next question comes from Clark Fuhs with Fulcrum Global Partners.

Hi.

Going back to the -- to the lower wafer-start growth.

It looks like with the higher yields, that the wafer-start ramp needed kind of slowed relative to your capital spending plan, which is -- which is net adding wafer starts, is that kind of a correct characterization?

I don't want to misinterpret what you said.

You said -- try it again, please.

Yeah.

The -- the wafer -- because of the higher yields, the wafer starts required to meet your projected demand are lower, so -- than you originally thought.

So the wafer-start growth is slowing.

Correct.

And the wafer-start capacity ramps that you're using with your capital spending is actually -- haven't changed yet, so you've got kind of a gap forming.

And that's actually what's causing the -- the higher cost absorption on a per-unit basis, is that kind of the correct way to think about it?

Not -- not really.

A lot of the second-half capital spend is on the next-generation, it's on starting to equip our factories, our shelves for 1264, which is the next generation beyond 90-nanometers.

Okay.

Of course we're still completing some 300-millimeter, 90-nanometer conversion, but I would think that the bulk of the spend in the second half is the next generation now.

Well, I mean, what you spent already is what's contributing to -- to the wafer-start capacity today obviously.

Yes.

All right.

Now, because of that lower wafer-start growth plan, I know Andy you present -- you always present this capital spending as a percent of cogs number, and you presented one in the May analyst meeting for 2005, that, you know, you could -- any -- any one of us with some assumptions could arrive at a range for capital spending.

Does that percentage now come down a little bit as a result of this slower wafer-start growth plan?

I'm not ready to update that one yet.

The -- I mean the simple answer would be if I -- cost of sales are higher, the math is going to make this chart look different just because of that.

The capital for this year's unchanged.

You know, you have to keep in mind that -- and our view today still, the capital needs, the capacity needs for next year are not changed.

We have a summer quarter/quarter and a half problem to get through which we will and as I said earlier, at that point we have to find a way to use these 300-millimeter factories, it's a big advantage for us and we -- we will find a way to employ them fully.

Okay.

Great.

And then my follow-up has to do with corporate demand.

It doesn't sound like you've -- you've seen any slowing in the corporate replacement cycle at any point in time?

No, we haven't.

Okay.

Great.

Thanks.

Your next question comes from Eric Gomberg with Thomas Weisel Partners.

Hi, good afternoon.

Maybe you could give us a sense of where you'd expect inventory turns to go, maybe by the end of the year, what -- what the target is given you're going to be working inventory lower.

I don't have a specific number for quote inventory turns.

What I said was I'd like to see microprocessor inventories down several hundred million dollars by the end of the year.

You know, one of the mistakes I made in the beginning of this quarter when I said I expected inventory to be flat, I didn't pick up the Flash success and recognize we needed to build inventory for the back half of the year.

So I think what you'll see is you'll see a real focus on getting microprocessors inventories down a few hundred million dollars, you will see Flash and chipset lower with inventories float whatever success we have taking share.

So if we are successful there, those inventories actually can go up in the back half of the year.

And just a follow-up.

Can you quantify how much of the -- of the margin guidance changes is due to mix with more lower margin products versus how much is due to lower wafer starts?

About half of the annual change is due to mix, the other half is due to microprocessor margins not increasing as much as I had hoped in the back half.

Okay.

Thank you.

Your next question comes from Kalpesh Kapadia with CE Unterberg.

Hi, Andy, Paul.

With a question on Asia Pacific.

That segment grew for you 11% in second quarter and driven primarily by China in both pieces and handsets and the Flash side grew 40% sequentially also driven by handsets.

From what we have seen, there's some slowdown in China handset market in May and June.

Is there impact embedded in your Q3 guidance?

Not per se.

And you have to really -- the bulk of our sales in Asia-Pacific are computer-related products, not Flash and handset-related products.

So that -- it's an overwhelming weighting towards the computer side.

And as I said earlier, I don't see any changes there going into the third quarter.

And have you seen any changes on the handset side from what they are used at?

In terms of consumption, it's hard -- it's hard for us to see that, you know, because we're in the mode of trying to regain share in some of these major accounts.

We've done a very good job at second- and third-tier manufacturers, which tend to manufacture in China, so that tends to be where some of our -- a lot of our Flash growth is but I've also said in my commentary that we did well in Europe and Japan.

All right.

And the follow-up to that on the PC side given the penetration in China is still low, do you think there's going to be any impact from the macroeconomic slowdown that is orchestrated by the government?

We haven't seen it and we were -- in fact the countertrend to that was the removal of the VAT, which actually makes PCs a little bit cheaper.

So you're seeing a demand continue to hold up in China?

Yes, we are.

Thanks.

Your next question comes from Mark Edelstone with Morgan Stanley.

Good afternoon.

Andy, question on the gross margin in Q2.

Can you give us an estimate as to what the benefit was to the quarter from the inventory build that you saw in the quarter?

Well, this is -- first let me just directly answer the question, which is not much.

And then, my -- trying to explain it is extremely difficult.

We're in a situation where what we sold in the second quarter was essentially what we built in the first quarter.

As a result of the build in the second quarter, it all goes into inventory.

If I had built half as many units in the second quarter, I just would have had half as many units valued twice as much, the inventory wouldn't have changed much.

Until I can shrink my days of inventory to be less than a quarter's worth, the real impact of those events take place a quarter later.

So what you'll see is the fact that we had good yields in Q2 - well, it actually helped me a little bit in Q3 which is why I said I expect most of the effect to be in Q4 of this slowdown.

Everything is lagging until I can get my inventories back down to less than a quarter's worth of inventories.

So not much effect of good yields on the second quarter.

Maybe just another way to think about it, then.

If you were not slowing your ramp, so if you had basically the same ramp in place now for Q3, what would be the difference in your gross margin expectations for the third quarter, all else considered equal?

Boy, that's a hard one to know the answer to.

You know, in general, take a -- again, take the full-year margin, which we took our midpoint estimate down by two points, half of that goes to the mix to the lower-margin products.

And so the remaining point, we're now in -- focused into the processor side of which there's -- you know, costs and a little bit of ASP pressure.

So take those -- you have to take those words, kind of do the math as to what a point's worth -- a little bit between the two and apply that back into the third and fourth quarter and it gives you an idea of what it would be.

Okay.

Just one last follow-up on the balance sheet.

You had a pretty big decline quarter to quarter in your deferred income in distribution.

Can you talk about that a little bit?

When -- a decline there is -- means basically we put less inventory into the distribution channel and nothing more than that.

So if you actually look at my -- just the inventory, they're down by, you know, $50, $60, $70 million.

My weeks of inventory are actually pretty low, I probably ought to have a little bit more inventory out there.

And is that equally split between processors and other products, or how does that mix look?

It's heavily weighted towards processors.

Great.

Thanks a lot.

Your next question comes from Nimal Vallipuram with DRKW.

Thanks.

First of all let me congratulate you on a great quarter.

Andy -- going back to this inventory issues, I hate to do this again, but --

That's fine.

If we take a step back, it does look like this Intel as a semiconductor company, is paying to some extent for doing your job better than what you would expect the -- you are paying for your yields being somewhat higher than expected.

So one would assume if that's the case, that your capital expenditure at least marginally should be lower going forward.

I'm not sure whether it's going to happen -- that could happen in the second half of this year or somewhere next year.

So that you don't get penalized for doing your job better than you were planning to do it.

Can you explain to me --

I agree with your analysis.

If we had realized the yields were going to be as solid as they were in the first half, we could have ramped our capacity at a different rate.

There would be less depreciation, there would be benefits.

That's why now that we have this capacity our goal becomes, try to make sure we use it fully and the way to do that of course are to get the processors all built there as quickly as possible.

If we still have capacity, they'll get chipsets out there, which may -- again may allow me to save some capital in the future, I understand and agree with that.

More efficient use of those factories should help in the long run.

Just a follow-up on that.

If you look at the inventory turns numbers, your current inventory turn is 4.1 time.

That is -- it has not come to this level, the last time it was as low as this was in third quarter of 1995, the -- that's about nine years ago.

Given, -- knowing what happened subsequent to that, do you believe that you can actually -- what are the chances of you bringing down the inventory to the level you want to be by the end of the year without having to take an inventory write-off in between?

I don't think I'll have to take an inventory write-off in between.

It -- I don't want to say it can't happen, but as I look at the current loadings, I look at the demand for the product, I feel actually fairly comfortable right now.

But I also don't think I get back to the lowest -- or the highest level of turns -- inventory turns that we saw a year or two ago by the end of the year, I think it takes a little longer than that.

I think I made progress this year.

Again, with the through-put times in my factories, the progress becomes noticeable in the fourth quarter and it needs to continue some next year as well.

Just finally, you didn't mention this.

If you can give us some idea without having to forecast the fourth quarter, that the improvement of inventory and subsequently the -- the work you are expecting to do to reduce the inventory, is it going to be mostly done in the third quarter or in the fourth quarter?

I really -- I can't make that -- I can't tell you that yet.

Based on -- it's really a function of what demand will be in the fourth quarter.

So we'll make that call not for another month or two.

Based on what I see today, the demand I would expect in the fourth quarter and the starts I'm making, obviously I think I'll eat up a fair amount of inventory in the fourth quarter.

If demand's better, I'll make more progress on the inventory.

If demand's worse then we face some of those other problems you were talking about.

All right.

Thanks, Andy, thanks, Paul.

Your next question comes from Krishna Shankar with JMP Securities.

Yes.

Can you talk about the rate of acceptance of the Grantsdale platform and the consumer markets versus the corporate markets and to what extent does the delay in the service factor from Microsoft impact Grantsdale's penetration into the corporate market?

Well, we kind of discussed the Grantsdale ramp earlier.

I don't really anticipate that the -- the effect of the one-week changed there where we were evaluating the product and doing the recall and swapping it out will have any material impact on the quarter or the ramp -- second-half ramp, rather.

The product is performance oriented so it appeals to both consumer and corporate, it is, it is tied to our -- what we call our staple corporate platform software so it becomes a very fast ramping corporate product as well.

Associated with Prescott die as that ramps.

I don't see any association with the delay of SP2.

Most corporations are on a -- you know, seat license and they gas pump the software as they want it not when Microsoft ships it.

Okay.

And to what extent can you stimulate elasticity in the PC market now that you have better yields and effectively higher processor capacity, what about being more aggressive on price cuts and trying to stimulate unit elasticity in the market.

There's no evidence that that works.

You know, we have a very broad-based product line, from, you know, the very high-end Pentium 4 Extreme Edition to the low end Celerons, and so there's -- there's product out there at virtually every price point today, so, you know, lowering the stack or changing the mix really does not change the demand curve.

Okay.

And my -- thank you.

Welcome.

Your next question comes from Tom Thornhill with UBS.

Paul and Andy, if you're looking at the yield issues where yield has been better than expected and I understand the loading issues are affecting gross margins, Andy, you've been -- you have put in a slide previously that talked about unit-cost reductions from -- on processors from Q4 of '03 to Q4 of '05.

Are we still on that track, excluding the loading issue?

Yes.

We -- I think I'll miss this Q4 only because the loading issue and we should be able to make it up in '05.

Making it up in '05 really is the question.

So once you get loading in line with -- and I understand you can't forecast '05 demand, but once you get loading back in line, should the efficiencies of 900 -- 90-nanometer, 300-millimeter come through in margins and cost?

Yes, that is -- at this point I would expect at this time.

There's nothing we've seen that says the formula's not right it's just that the loadings are not as high as they could have been.

So then the long-term forecast for -- longer term forecast for gross margin is impacted by the strategy of moving up motherboard, Flash, and chipset, which I think -- here's the question.

Should we presume that that's a permanent change in strategy part of which is loading the back -- the back end -- not the back-end capacity, but the 130-nanometer capacity, but then other than that issue on -- on strat -- on margins, we ought to see the processor margins kick back in when the loadings get back in line.

Yes, I believe all that is correct.

You know, the -- it -- the strategy to expand -- obviously, we've been trying to take back the Flash space for a long time.

It's good to see us being successful there and there's no question the team Craig and Paul have in place are going to drive that one real hard.

The chipsets, -- it's not a opportunistic, it's actually a long-term goal to increase our tax rate, but we will go actually that one, motherboards are the same.

So the answer to the question is yes, that is not opportunistic, it is a long-term strategy.

Thank you.

Your next question comes from Joseph Osha with Merrill Lynch.

Hi, guys, just back to the demand modeling perspective here.

I know, Andy, you've been staying sort of deep all year, but I just want to clarify this.

Something has to have changed.

In -- you know, your assessment of what the demand environment looked like for the second half, is that true?

Not really.

If we shared with you planned units for processors, put together in December and I shared it with you today, it would be remarkably stable.

Probably what changes, Joe, is as we begin the year, in our business, you want to make sure you don't miss a unit, you make sure you have capacity for more overbuild.

As you get halfway through the year, you don't need -- by overbuild, I mean overdemand.

As you get halfway through the year, you can start to narrow your margins on top of what you expect to happen, so that you don't have to have quite as much plan to build extra inventory.

In reality, our view of the year in the microprocessor space is amazingly constant.

But I guess the implication of that, then, is I've seen -- been following you guys for a while and I've seen you go through, you know, various different technology ramps and I've never seen you misjudge your output this -- this massively.

Are you really -- really saying that, you know, the process -- that all of this run up in inventory on the microprocessor side has simply come from, you know, the more -- more stuff coming out the end of the pipe than you thought?

Yes, actually, it's a -- if you -- if you can think back to January and February, when you guys were writing articles that said they can't yield Prescott's, 300-millimeter's in trouble.

Actually I wasn't, I was writing you had too much inventory, but go ahead.

Good.

I'm glad you got it right.

But anyway, there was a concern in our shop as well, you know, this 300-millimeter, 90-nanometer stuff is new, I think we're probably one of the few who are actually making it work and we were very concerned at the time.

At this point, I would tell you the process works wonderfully, it's healthy, and it's one of the good surprises you then have to figure out how to deal with.

A high-class problem to have.

I guess the second question, then, and perhaps you can just comment on my logic, is that as we look at the 65-nanometer and the plant spend there, you know, one presumes that you're sort of model for capital productivity is going to change, right?

The -- yes.

I would expect in each succeeding generation to get a little less productive, a little less efficient, let's put it that way.

I'm sorry.

Capital efficiency does suffer with each new technology, yes.

But is that changed relative to what you would have thought given how well you've done on guidance?

No.

It's another one of those things where you'll always be cautious.

What you can't afford to have is your leading-edge best product come up short because of a ramp situation.

So what you'll tend to do is go ahead and plan the beginning of your ramp the same and then modulate how you deploy the next second and third facilities.

Okay.

Great.

And then one just sort of off the range question.

You know, as you look at some of the 130-nanometer, 200-millimeter capacity, what kind of discussions are going on inside Intel about perhaps other areas that you might move into that, you know, are of interest that need to load the stuff up?

Well, I mean, obviously, the chipset business is the prime capacity driver for that.

And there's a lot of opportunities in the general areas of peripherals that can be integrated into that chipset.

Graphics, communications, IO kinds of devices are the most obvious.

You can spend more of those transistors on things like graphics, you can spend more transistors on multiprotocol radios and those kind of things.

Everything we've been talking about in the chipset road map is available for construction on that level of psychology.

Would something like freestanding graphics, processors ever get back on the table?

Not really.

At least not in the foreseeable future.

Because to compete there you really need a generation M, not M -1 and the current -- the high-M capacity is much better served by building microprocessors.

Okay.

Thank you very much.

And we're worried about the overall thermal envelope of these PCs anyway and, you know, helping that -- by taking a 2 or 3 million transistor graphics product will not help us.

Thank you very much.

Okay.

Operator, we'll take two more questions, please.

Okay, sir, your next question comes from David Wu with Wedbush Morgan Securities.

Good afternoon.

Can you talk a little bit about the Flash memory business?

On the -- I guess there's -- the Nan Flash which is one of the competing technologies coming out pretty hard, potentially even harder in the second half, I was wondering the improvement that we have seen to date in the Nor Flash business of yours, could that be any jeopardy where there's some cross-price elasticity involved here?

The other thing I was wondering is, Andy, you'd mentioned in the last conference call that the Flash memory because of the relatively low -- low loading, any improvement in Flash really translates to pretty good gross margins, despite the fact that Flash is normally loaded in microprocessor.

Are those economics still at work or are we past that stage where you get very high incremental gross margin in these Flash memory business, and one quick question on the motherboard business.

I guess -- I understand why you want to ramp Flash and want to ramp chipset on -- on 8-inch wafers.

What is the rationale for ramping your market share in the motherboard business?

I'm not sure that that's ever very profitable for anybody and even guys in Taiwan don't make a lot of money.

I'll do the third, if you do the first two.

Okay.

Let me start with Flash loadings.

That's the quickest and easiest.

With the rate of growth we saw in the second quarter at 40% quarter to quarter --

Yep.

Sadly, you see all the gifts you get from the initial ramping.

So I still get more margin, I get positive gross margin out of growth in the future, but it won't be at the same rate I got in the second quarter, so if you look at the second quarter where, you know, we improved the bottom line by almost $100 million, the same type of revenue increase in the future won't give me quite as much of the bottom line.

I see.

As far as competing against Nan, there's no question there's going to continually be a tug of war as the Nan companies, particularly the big companies are trying to take that into the Nor space, you'll see us trying to make the Nor competitive to compete against the Nan space.

I think you will continue to see skirmishes along that line.

I don't think you will see any right turns, major changes overnight.

As we watch it over the next, you know, four to six quarters it will be interesting to see what kind -- if there is a permanent change, and if there is how much.

At this point, I don't think it affects the rest of this year, and I think it's something we have to watch for next year.

Okay.

Fine.

What about motherboards?

Okay, David it's Paul.

Hi, Paul.

Yeah, hi.

There's really two fundamental reasons that we're -- that we have remained very aggressive in the motherboard business.

One is that it -- we have found over almost two decades now, that it is by far the best vehicle for us to be able to drive new technologies and technology transitions through the marketplace very, very rapidly, to have all of this debugged at the board level is very convenient for our customers, not just OEMs but also channel customers.

Uh-huh.

And gives us that acceleration factor number one.

Number two, the motherboard business isn't just a desktop business.

A very large number of our server chips are sold associated with our server motherboards, and that allows us to gain the channels that participate in the segment of the market that they enjoy because it's more profitable and we like to support them.

I look at this as being a very strategic business.

Paul, do you -- do you have a target in mind, let's say, 30% or 50% of the total motherboard business out there?

Oh, not one that I'm particularly willing to share on this call.

Ha-ha-ha.

Okay.

Okay.

So I got to average -- average your margins between motherboards and chipsets and server microprocessors, that's the way you have to look at it, right?

Blend is a better word.

Okay.

Fine.

Thank you.

Your next question comes from Graham Tanaka with Tanaka Capital Management.

Yeah.

I wonder if you can explain a little more about this shift and if it's a pretty significant year to year from U.S. to Asia and what your estimate is and what percentage of revenues have -- have sort of transferred over to Asia, to understand the growth rate, you know, 31% in Asia even flat in U.S.

Yeah.

It -- Graham, that's the million-dollar question, right.

Right.

We have estimates of that.

I'm not terribly comfortable showing.

We estimate consumption, PC consumption in every country on earth.

Right.

And then we -- for APAC, what we do is we back out the amount that we ship in versus the amount that we think is consumed there and the rest is for re-export back to Europe or -- or -- nowadays in Japan and the United States as well.

That's why -- the best indicator I can give you is the one I put in my commentary, which is that we still see strong year-over-year growth in consumer and corporate in the United States, which is not showing up in the revenue numbers as they are broken up by geography for us.

Well, the -- the -- can you give us an estimate roughly of what kind of growth rates they were?

I mean many of the analysts are struggling to look at retail sales and use PCs in the U.S. and use that as a guide for Intel's overall-demand and I think that's --

I'm not comfortable with us winging it.

I mean, I'd rather do some thinking and get some data and get back to you that way.

Thank you.

Okay.

Okay.

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

A recorded playback of this call will be available at approximately 5 p.m.

Pacific time tonight.

Those interested should dial 1-800-642-1687 and reference pass code 8019927.

Thank you and good night.

This concludes today's conference.

You may now disconnect.