英特爾 (INTC) 2006 Q3 法說會逐字稿

Good day ladies and gentlemen, and welcome to the Intel Corporation Q3 2006 earnings conference call. [OPERATOR INSTRUCTIONS] It is now my pleasure to turn the presentation over to your host for today's conference, Mr. Alex Lenke, Investor Relations Manager.

Please proceed, sir.

Welcome to the Intel third quarter earnings conference call.

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

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

The third 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 uncertainty 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 cost.

Of any mergers, acquisitions, divestitures, investments, or other business combinations that may be completed after October 17, 2006.

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

With that, let me turn it over to Paul.

Thanks, Alex.

Q3 was a good quarter for Intel across many fronts.

Sales were above the midpoint of our expectation in July.

We delivered over 6 million processors based on the new Core microarchitecture and had record shipments in our mobile and server product segments.

Chip sets remained strong on a solid note and we gained market segment share.

In the past 100 days, we have introduced a wide array of new products, refreshing our entire product line from notebook to desktop and from volume DP servers to Xeon, MP and Itanium.

We are pleased the industry recognizes our clear leadership with the Core microarchitecture.

And we're disclosing today that we have now shipped quad-core versions for revenue, as well.

Intel's quad-core processors work in existing designs, boosting energy performance by up to 50% in DP servers and by about 70% in desktops and workstations.

Our factories have been executing extremely well with over 40 million 65 nanometer processors now shipped and a crossover from 90 nanometer now achieved.

Our 45 nanometer process development is on track for the second half of 2007 and we plan to tape out the first of 15 45 nanometer processors this quarter.

Our R&D strategy is to introduce a new microarchitecture every two years, a pace that we haven't seen in a long time, while shrinking the existing architectures in the intervening years for an unrivaled innovation stream.

Moving to the digital enterprise.

Our server business generated higher revenue in record units driven by strong demand for new products in Xeon DP, MP and Itanium.

We launched the vPro platform for business PC's and see strong design wins and ecosystem support.

The mobility group also had record units as mobile continues to take client PC share.

We demonstrated our new [Ropeson] technology, which places NAND flash on the mother board for faster boot and resume, plus longer battery life.

We introduced the industry's first WiMAX silicone for both fixed and mobile networks and see WiMax as a potential catalyst for both Centrino and our ultra-mobile PC platforms.

The desktop market segment remained very competitive.

We are pleased with the market's acceptance of our moving the Pentium product into lower price points, as we ramp the new Core microarchitecture.

The Core 2 Duo is now recognized as the performance and energy efficiency leader.

And is moving into the mainstream supported by our biggest ad campaign in years.

At the spring analyst meeting, I announced a comprehensive review of the Company's operations with the goal of making Intel a more nimble and efficient competitor.

During the quarter, we reached a number of important conclusions and also began to take significant actions to reduce our ongoing costs.

Our head count declined by over 2,600 during the quarter and we are on track to be at 95,000 employees by the end of the year and 92,000 by the middle of next year.

We plan to reduce our costs by $2 billion in 2007 and by $3 billion in 2008.

In summary, Intel is entering the strongest part of the year with technology and manufacturing leadership, ample leading edge capacity, and a portfolio of new products across our mobile desktop and server businesses.

We plan to sustain our lead by being the first at 45 nanometer and delivering a new core or shrink every year.

We will achieve these goals by becoming more agile and putting the right cost structure in place for our future growth.

With that, let me turn the call over to Andy.

Thanks, Paul.

We made good progress in the third quarter toward our goals for a stronger second half.

Revenue from microprocessors was up from the second quarter in all geographies and in the channel.

The server business was an important bright spot with double digit sequential and year to year growth.

Our comprehensive effort to reduce costs is gaining traction, as is our work to restructure the business.

Gains from divestitures and sales of investments contributed to earnings per share.

Revenue for the third quarter was $8.7 billion, in the upper half of the range we forecast in July and up 9% from the second quarter.

While consistent with patterns for this period, a portion of the sequential growth is a function of our lower revenue base in the second quarter when some of our customers reduced their inventories.

The growth in the third quarter came from higher unit volumes of microprocessors, chipsets, motherboards, and wireless networking products.

Offset somewhat by lower average selling prices for microprocessors.

The server business led the percentage growth from the second quarter with higher units and higher average selling prices.

Microprocessor revenue in the digital enterprise and mobility groups grew 9% to a total of $5.8 billion.

Revenue from chipsets, motherboards, and other products in these two groups was $2.2 billion, up 11%.

Flash memory was lower than the second quarter due to lower unit volumes.

As is usually the case in the third quarter, business was strongest in the greater European region where revenue was up 17%.

In Europe, the Americas and Japan revenue growth was higher than seasonal averages, while in Asia Pacific the growth was slightly lower.

In a year to year comparison, quarterly revenue was down approximately 12%.

Most of the decrease came from the digital enterprise group whose revenue was down approximately 22%.

Revenue for the mobility group was up 3%.

Among the geographies, nearly all of the year to year revenue decrease was in Asia Pacific and Europe.

Gross margin dollars of $4.3 billion, including the effect of share based compensation of $103 million, were 3% higher than in the second quarter.

Gross margin percentage was 49.1% on a GAAP basis, in line with our forecasts.

Without share-based compensation, gross margin percentage was 50.3%, 2.6 points lower in the second quarter.

Some of erosion in average selling prices, higher unit costs for 90 nanometer duel-core microprocessors and reserves for mature products lowered the gross margin percentage, while higher unit volumes partially offset the decrease.

In a year-to-year comparison, gross margin percentage, excluding share-based compensation is approximately 9.5 points lower for the first quarter of 2005, primarily a result of lower revenue.

We exceeded our goals in reducing spending during the quarter as we lowered head count and cut discretionary spending.

Spending is also down sequentially as a percentage of revenue.

R&D and MG&A were $2.8 billion lower than the forecast in July and down 9% from the second quarter.

Excluding share-based compensation of $232 million, spending decreased approximately 8% from a year ago.

The number of employees declined during the quarter to just below 100,000.

In addition, restructuring charges for the quarter, consisting of severance and benefits for terminated employees, were $98 million.

Net gains on equity securities and interest in other were $440 million, two times the $220 million forecast in July due to gains from divestitures and the sale of a portion of investment in Micron Technology.

Fully diluted earnings per share were $0.22, including approximately $0.015 from the net impact of gains from divestitures and sale of investments and from restructuring charges.

Excluding share-based compensation, earnings per share is $0.27.

On the balance sheet, inventories of $4.5 billion were up $145 million for the second quarter.

Inventories of chipsets and flash memory were higher.

And inventories of microprocessors were lower.

Based on the midpoint of forecast for revenue and gross margin for the fourth quarter, inventories at the current level are less than one quarter's cost of sales, which is similar to our throughput time.

We expect inventories to be slightly lower by the end of the fourth quarter.

Total cash investments, comprised of cash, short-term investments, and fixed income trading assets, ended the quarter at $7.8 billion, up from $7.2 billion in the second quarter.

Capital spending was $1.2 billion.

Stock repurchases were $500 million.

And dividend payments were nearly $600 million.

As a reminder, we indicated in April that we expected to reduce the level of our stock repurchases as we continue to manage our cash flow.

As we turn now to the outlook for the fourth quarter, please keep in mind that unless otherwise specified, the forecasts do not include the affect of any new acquisitions, divestitures, or similar transactions that may be completed after October 16.

With the exception of the head count forecast, the outlook does not include the effect of a previously announced planned sale of Intel's communications and applications processor business to Marvell Technology.

I will use the midpoint of forecast ranges when making comparisons to specific periods.

We are planning for revenue in the fourth quarter to be between $9.1 and $9.7 billion and an increase of approximately 8%, consistent with seasonal patterns.

This forecast assumes higher revenue than the third quarter from microprocessors, chipsets, and flash memory.

Our expectation for gross margin percentage in the fourth quarter is 50%, plus or minus a couple of points.

Without share-based compensation of approximately 1 point of margin, the forecast for gross margin percentage is 51%, plus or minus a couple of points.

Higher volume and slightly lower unit costs from microprocessors will lift gross margin, but we expect this to be partially offset by charges on underloading 90 nanometer networks in the factories, as we transition production to 65 nanometer and by start-up costs on the new 45 nanometer process technology.

Spending, R&D plus MG&A, should be approximately $2.7 to $2.8 billion.

The forecast includes about $250 million of share-based compensation.

Without the impact of these charges, spending should be approximately $2.5 to $2.6 billion.

Although revenue is forecast to be 8% higher than in the third quarter, we expect spending to be approximately flat or slightly lower as we proceed with cost reductions.

In addition, we expect restructuring charges of approximately $125 million.

We are on track with the previously announced plan to reduce head count to approximately 95,000 by year end, including the effect of the planned sale of the communications and applications processor business.

This would bring head count down 5% from the end of 2005, with continuing reductions through 2007.

As a result of several cost savings actions, we have lowered by $100 million the forecast for R&D for the full year.

Now set at approximately $5.9 billion.

Without share-based compensation of approximately $500 million, the forecast for R&D is $5.4 billion, up 5% from 2005.

We also continue to find savings in capital spending and have again lowered the forecast for the full year by $400 million to $5.8 billion, plus or minus 100 million.

The largest portion of the savings comes from initiatives directed towards improvements and productivity.

The estimated tax rate for the fourth quarter is approximately 30%.

In summary, outstanding new products, leadership and manufacturing technology, and disciplined execution point to a stronger second half with seasonal growth as we lay the foundation for 2007.

While business remains comparative, we remain focused on doing the right things.

Delivering compelling products, competing for sales, attacking costs, executing superbly with 65 nanometer and pushing ahead with the next generation.

With that, let me turn it back to Alex for Q&A.

Thanks, Andy.

We will 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 INSTRUCTIONS] And sir, our first question is from David Wong.

Thank you very much.

Can you just clarify your gross margin guidance for the December quarter?

Does it assume any inventory charges?

And in line with this, can you also give us some feel for your current inventory?

How much is older microprocessor product, say 90 nanometer microprocessor product?

David, the forecast for the fourth quarter for gross margin percentage, as I said, we expect to get some benefits from higher revenue.

We expect to get some benefit from slightly lower cost per units.

And then that's offset by -- I actually mentioned two but there's three factors.

The one factor was some underload charges in the 90 nanometer network, start-up charges in the 45 nanometer.

And quite frankly, the third piece that I didn't was the fact that I did take some reserves in Q3.

I don't plan to re-take reserves in Q4.

So when you do a quarter comparison, that would show as a positive.

So no, there's no unusual reserves.

There's always ins and outs of inventory but it's kind of flat and normal in the fourth quarter.

In terms of how much is old versus new, we've made the transition, as Paul said, into the new product.

If you notice in our inventory, WIP is a huge number.

That's all new products.

So, the inventory is transitioning pretty rapidly into the 65 nanometer product line.

Great, thank you.

And sir, our next question is from the line of Joe Osha with Merrill Lynch.

Wow, I made it early this time.

Can you give us the sense, again, looking at sort of the CONRO versus Presler mix when that toggle will be complete when you'll no longer expect to be shipping the older parts?

Well, I think we'd like to like to reserve the right, Joe, to use Presler as a dye down -- in the low end of Pentium and even the value segment over the course of next year.

As we -- we'll have a number of options down there.

But they'll all be 65 nanometer based.

So let me rephrase that, then.

When do you expect to have the mix in your performance desktop to where you want it to be?

Well, a lot of that has to do with how fast the market moves to Core 2 Duo and that is one we're ramping as fast as we know how to do.

As I've talked about, I gave you the numbers at IDF and again today.

We gave you an additional point on that today, which was we talked about the totality of shipments of Core microarchitecture, which includes Woodcrest.

And Woodcrest had over 1 million units in its first 90 days.

That one product alone has now achieved over 40% of the entire DP market in its three months of its existence.

So, we're very happy with the ramp and we'll continue to push it real hard in Q4 and Q1.

Okay.

Thanks, and just a housekeeping question for Andy.

Can you give us some sense as to -- as we look into 2007 how we might think about the expense run rate?

Is it -- might it sort of -- can we take the Q4 run rate and just map it through '07 or is it going to drop further or go up?

Any guidance there would be appreciated.

It's really way too early to give you a definitive number.

That said, of course, I'll stick my neck out a little bit.

I sure would hate to be above Q4 and if I can find a way to be a little bit below it, I would.

I still think we have room to improve in 2007.

Okay.

Thank you.

And sir, our next question is from the line of Glen Yeung with Citigroup.

Andy, this is actually just a follow-up to Joe's last question.

If I go back 10 years or so ago and we looked at the OpEx as a percentage of sales, you were talking maybe 20% then versus close to 30% now.

Is 20% over a longer term period an achievable target?

I don't want to give us spending as a percent of every target for pir long-term.

What I would say is, I think, as I said, I think Q4 is a high, not a low.

I think we need to let revenue grow and maintain that.

So whatever you think revenue is next year, take flat to slightly down through the year for spending and then we'll talk next year about how far we can drive it beyond that.

The other point I wanted to question you about is inventory.

I think you said it would be slightly down in the fourth quarter.

And given how high it is, it's a little surprising.

I wonder -- you're actively working it down, I would assume because you're taking some underutilization charges.

Why not down more?

And is there a level to what you're trying to manage these inventories levels that may be different than what we've seen in the past?

It'll be different at different times.

As we've done the transition, the Broadwater chipset into the 90 nanometer factories as we begin to tee up for the Broadwater equivalent for the mobile space.

As you ramp the Core 2 Duo products into the 65 nanometer, you can't pull back on those.

That is your future, those are the products that will win in the marketplace.

So, what you tend to do is drive the new technology, the new products into healthy inventory positions and then you manage the older products down slowly and carefully to make sure you don't get caught without the ability to meet a customer demand.

So, so I guess it's a long way of saying, yes, I think at today's revenue level, 4 or 5, a little higher than I would like but through a transition, it's not a bad place to be.

Does that imply that we need to see future underutilization charges?

Just as you sort of balance that out if you're a little bit overbuilding now, I assume that at some point you get a little bit of underbuild just as things start to smooth out a little bit?

I don't want to make a factory loading answer for next year before we've given you any insight as to our thinking for next year.

In general, what I'd tell you is we have looked at the factory network for next year.

What I said, was the underloading is in the 90 nanometer 12-inch factory.

Some of which we plan to convert to leading edge technology.

Some of which we plan to fill up, like I said, with the 90 nanometer Broadwater equivalent for mobile.

So, I looked at the network for next year, and I don't think I have a big problem with excess capacity.

It would take a pretty noticeable change in the marketplace for me to be concerned about that.

Good to hear, thanks.

And sir, our next question is from the line of Sumit Dhanda with Bank of America Securities.

Yes, hi Andy and Paul.

A couple of questions.

First off, as it relates to the fourth quarter outlook for margins, could you help quantify how much benefit you're seeing from lower unit costs?

I think at your analyst day you had suggested it would be somewhere 5% and 10% based on a chart that you had put up.

Whether that still holds true?

And then, what's the exact level of offset from the underloading charges that you're seeing on 90 nanometers and the 45 nanometers start-up costs?

And then I have a quick follow-up.

So, I'll give you partial answers.

The underload charge and the 45 nanometer start-up charge are equal to about 2 points of margin, combined.

The lower cost units are providing a small benefit.

I haven't looked and compared it to the chart we showed.

My guess it's in the neighborhood, maybe not exactly there.

And the other lift is the fact that revenue, if we hit our midpoint, will be up a pretty decent 8%.

Okay.

And I guess the second question I had as it relates, I know you said -- you didn't quite give an exact answer to Joe's question, but the $2 billion cost savings that you're referring to, is there a benchmark that we could use in terms of what number are you really comparing that to?

Is your expectations back earlier in '06 for '07?

How do we think about that?

I'm -- simply going to ask for your patience until January to give you a definitive answer.

What it's compared to is when we started the program.

So, when you think of how to model next year, we have some of the savings in this year's run rates.

And what we'll do is -- so starting from this point, we should -- starting from that point say, arbitrarily pick a month, pick May or June, we would actually have the run rate be at $2 billion lower than the run rate was at that point in time.

So, you've seen Q3 be a little bit lower, you'll see Q4 be a little bit lower.

That gives you the ability to at least kind of put a point on a map and get close.

And just one follow-up here.

The Asia Pacific sales down double digits year-over-year, so much growth in those markets occuring.

What do you think is the reason why you're not seeing better growth in that market at least from a year-over-year comp basis?

Is just the inventory overhand you have to work through?

That's a bit of it.

There's also been -- the Asia Pacific market is a much larger percent of PC's are desktops versus notebooks.

The mix is more of the desktop flavor.

And desktops always have the most acute price degradation.

So I think that that's probably been the driving factor there more than anything else.

The other thing that happens in Asia is that the bulk of the notebooks of the world are built there and then trans-shipped back out to customers around the world.

Okay.

Thank you.

And sir, our next question is from the line of [Ousha Orji] with UBS Securities.

Thank you very much.

I just have two questions.

If I look at your guidance for next quarter, how shall I think ASP units standard mix, given that we had slightly lower ASP for Q3, which it seems units might have been up at least 9% for Q3.

And looking into Q4, how should we think about that mix and at what point should we expect the mix of new products to kind of kick in and help stabilize ASP's?

So I think the new products are already helping.

The price pressure we saw in the third quarter was less than we saw in the second quarter.

I'm not going to provide an ASP forecast for the fourth quarter but again, we're seeing so some level of firming based on the new products.

So, we'll always be in an environment where it's competitive.

We'll always be in the environment where price is one of the factors in the marketplaces.

All I can tell you is Q4 was worse than Q3, which I think will be worse than -- I'm sorry Q2 was worse than Q3, which I think will be worse than Q4.

That's great.

Just, a quick one on flash.

Given that NAND flash is going to ramp up at any moment, it seems to me like NOR flash may have had a significant weakness.

Are you able to provide some color as to what happened there?

Well, let me take that one.

We have -- most of our business is still primarily large customer based.

And -- coming off of the shortages last year, we were not able to move into the channel as rapidly as we would have liked in terms of expanding our product line.

And so from my perspective, that probably cost us a little bit of share offset by slightly better pricing scenario than we would have had otherwise in NOR.

We are looking for NOR to grow, as Andy said, in the fourth quarter.

That's helpful, thank you.

And sir, our next question is from the line of John Lau with Jefferies and Company.

It seems that pricing seems to have firmed up.

And I was wondering, what are you seeing in the marketplace in terms of the end markets?

Are you seeing solid seasonal demand?

Is that why the pricing environment is now more benign?

And is a lot of the chipsets you're putting onto the marketplace that are being absorbed quickly because of these end markets are solid?

I think it's a bit too soon to characterize the market for the fourth quarter, we're just now getting into that.

But it looks very good, it looks seasonal, the guidance you're getting from Intel.

In terms of the pricing, I think it's more reflection of two things.

It is one, you saw Intel reset its entire price stack top to bottom in the last three or four months, which made room at the top for Core 2 Duo.

It pushed Pentium down and it pushed Celeron down.

Giving us a three tier price stack for the first time.

That was a rather large price move in the grand scheme of things, perhaps the largest we've ever made.

And as that played through and the mix started shifting in there, that's the overwhelming dynamic.

The second one is how fast the Core 2 Duo ramp, particularly in servers and notebooks where pricing tends to be a little bit better than desktops.

And then finally, you mentioned that you're doing quite well in terms of growth in the service space.

What would you attribute the largest reason for that?

Is it the Woodcrest that had come out in terms of their performance, power, or the price?

I think it's a set of very good products around Woodcrest for DP and Tulsa for MP, the not only had great price performance metrics, but also had perhaps the most significant scale we've ever had in terms of number of customers, number of SKU's lined up.

And that was one of the things I think you saw, if you went to IDF, you saw customers that had not been with Intel in the last couple of years returning back to Intel very strongly in a lot of the Web 2.0 and RAC kinds of implementations.

The third point I'll point out, is that we actually had our all time record quarter for Itanium in the last quarter.

And that really was driven by Montecito and Itanium continues to be a very good product.

It gained share and is gaining share against spaRk and Power at the high end.

So all three of those product lines were beneficial to the overall results in Q3.

Your service base has been quite competitive.

And so with the success that you're having and the traction that you're gaining in this marketplace, do you know how much market share you may have gained back in the Q3 time frame?

I think it's too soon to call that level of granularity.

In fact, the people that track this stuff don't even publish until four months after the quarter, or so.

You won't see this one publicly for a while.

But as we said, our view is that we did -- by setting all-time record in server units, we probably gained some share there on a billings basis.

That's great, thank you very much.

And sir our next question is from the line of Jim Covello of Goldman Sachs.

Good afternoon, guys.

Thanks so much.

First question kind of goes back to the pricing issue.

I understand pricing has gotten a little bit better over the last couple of quarters, kind of quarter-over-quarter.

The question is, as you guys continue to gain regain share and your competitor continues to add significant capacity, is there any risk that the competitor gets a little bit over aggressive on pricing and that constrains the kind of margin expansion we expect for next year?

How do you think about that?

You'll have to ask them what they plan to do on pricing.

Let me just address the capacity thing from a -- not all capacity is equal, Jim.

And we are -- well over half of our microprocessors are on 65 nanometers now, we're headed towards 100%.

Independent of whatever the environment is next year, it's my opinion, having seen this for a few decades, that the leading edge capacity and the best products always sell off first.

I think Intel is an extraordinarily good position, with both products and capacity coming on and a transition to 45 nanometers coming next year, to do well regardless of the profile of the market.

So to paraphrase that, the better technology helps you to avoid a real price war like you might have seen over the last 18 months or something along those?

Price wars tend to happen in the brown banana segments.

Final question for me, on the NOR flash segment, any potential for incremental restructuring in that segment as that maybe lags in profitability a little bit?

Well, the operating income for the business unit is actually improving a bit.

You can't really see that the way we break out of the numbers because the NAND numbers are combined with the NOR numbers.

But you may remember that a couple of quarters ago we shipped the NOR manufacturing and technology development teams over into the size of the NOR business and unleashed them to be able to run any way they need to run to be competitive.

And we're starting to see the benefits of that shift playing out in our operating expenses and in our factory expenses as we go through the successive quarters.

Thank you.

And our next question is from the line of Tim Luke with Lehman Brothers.

Thanks.

Paul, I was wondering if you could give us any color on your perception of the shared dynamics in maybe the desktop and mobility area?

And how given your guidance you anticipate that going forward into the end of the year?

Well, all I can do is reiterate what we've already said here, Tim.

It was a record, we believe we gained some overall share this quarter.

We had record unit shipments in units and notebooks, which are the two fastest growing segments of the market and desktop is still a very, very competitive area. vPro, which launched in the quarter, is not yet in very high volume at many manufacturers.

So you start seeing the impact of that with Core 2 based products later this year in the fourth quarter and first quarter of next year.

In terms of the overall share dynamics, we said in April our goal is to gain share throughout this year, that remains our goal.

Okay.

And with respect -- as a follow-up.

As you move into the year-end quarter, via normal seasonality perhaps the benefit of adding a new customer that's reasonably in the second half of this year, balanced perhaps by the timeline associated with the Vista upgrade.

Could you talk a little bit about how either of those factors may be helping you or otherwise adversely impacting you for the year end?

I certainly can't give you any insight into our [Coopertino] based customer because it would be inappropriate.

But so far they've done very well with the transition.

And I think that the momentum that they had coming into this quarter was really quite good for them and this is their back-to-school quarter.

So, you'll have to wait for their results to come out to talk to them.

But I think you'll see -- you'll be happy there.

Vista, just your perceptions of that?

Yes, I just got back from the Gartner conference.

And from an IT perspective, I think Vista is not something that people are exactly waiting for.

And there's going to be a measured deployment of this as near as I can tell.

At least that was the Gartner view and the CIO view that I saw down there.

I do think that on the other hand, though, you'll see it be instantaneously adopted, particularly in new machines, in the consumer market.

And so it's probably a lot like we saw with XP where you see a very large chunk of the market, small and medium business, consumer move quickly and enterprise move more slowly.

Does that dampen the fourth quarter then or help it in your opinion?

I don't believe so.

One of you guys asked me this question last quarter.

And in the range of the last five years of seasonality, we haven't had a new operating system.

So when we talk about seasonality, we're really reflecting the norm without the new one versus what would happen if one has shipped this quarter versus next quarter.

All right, thanks very much.

And, sir our next question is from the line of Adam Parker.

Yes, hi, just a couple questions.

You keep saying that the desktop market is competitive, Paul.

And you talked about how leading edge capacity and best products sell first and how you gained some market segment share.

Can you talk about the uptick of CONRO?

Has that been what you had anticipated?

Has it been disappointing at all and what are some of the variables there?

The overall uptick of Core 2 Duo in PC space has been exactly where we thought it would be.

Merome has been a little bit better than we thought but CONRO was a little bit looser than we thought.

And the principal reasons are infrastructure in both cases.

Merome plugged into an existing Yonah socket.

And CONRO , in particular for the high volume SKU's out of the larger multi-nationals, awaited the integrated graphics version of Broadwater, which didn't really ship until September.

And so the volume was a little bit later in terms of the high volume ramp of that one.

The gamers all came out with it and the high-end discrete graphics machines were all out there but the integrated graphics volume points were a little later in the curve than we first thought.

I think we're beyond that point now.

All of the majors are now shipping with integrated graphics and multiple price points.

And now it's just a matter of ramping into the very high volume price points over the next couple of quarters.

Okay.

And a follow-up is just, Paul, how do you think raising pricing or improving the pricing environment versus ceding market share?

In other words, which variable is more important to you?

Because the obvious math says raising pricing is more important and it seems like you're considering duopoly pricing environment but you seem also focused on market share.

And I'm trying to figure how you're balancing those two things?

I'm very much believer that the best technology wins.

Right now Core 2 Duo, CONRO, Merome and Woodcrest are the best products.

The fast we ramp those, the better chance we have of retaking market segment share.

We use the Pentium brand as a very effective tool in the channel in particular to retake share there because it's a healthy product with a good brand name, good recognition, very attractively priced.

I don't have a recipe I'm going to give you for share versus pricing because it -- the market doesn't really work that way.

I know that it's very important to us in a high capital business to make sure that we maintain and continue to grow our historic share.

Okay so that's -- all right.

Thanks.

And sir, our next question is from the line of Ross Seymore with Deutsche Bank.

Thanks, guys.

Just a couple questions on the pricing environment, trying not to be too repetitive with other people that asked.

Andy, in the past you've given us a little bit of breakdown in the ASP pressure, whether it was coming from mix or like for like price cuts.

In this past quarter, can you give us an idea of how that blended together?

Not, I can't give you much on it.

We told you service did well.

We told you Core 2 Duo is doing well.

Those things are helping.

Desktop continues to be the place where you have higher price pressures.

So, it's not much difference.

You also realize, it wasn't a major factor in the Q3 margin percentage drop.

It was just one of three and probably equally weighted.

And in the fourth quarter, I didn't mention ASP as a factor for margin change.

So going forward, does there come a point that your mix -- well, the mix component of ASP's becomes a tail wind rather than a head wind?

Again, it's tough to speculate on what ASP's will do next year.

There are a lot of different variables.

For right now, I want to just leave it.

I think the Q4 environment is a little better than Q3 environment.

And then one final one on the inventory side of the equation.

Last quarter you talked a bit about the valuation creep as the new products move from raw materials to WIP.

And you mentioned that this quarter, as well.

In your guidance for inventory to drop slightly in the out quarter, how should we think about units versus that sort of just markup as the new products move to finished goods?

Boy, you'll get a very different answer across different parts of the product line.

So, if you said 65 nanometer processors, units will go up.

If you want 90 nanometer processors, unit wills go down pretty markedly.

If you go 90 nanometer chipsets, it will go up.

If you go older technology chipsets, it will go down.

So in total and overall, it probably won't change a whole lot, but the quality of the mix will improve pretty dramatically.

Okay, thank you.

And sir, our next question is from the line of Mark Edelstone with Morgan Stanley.

Good afternoon, guys.

First question for you, Paul, if I could.

I think your microprocessor units overall were at record levels in the fourth quarter of last year.

We've had down comparisons as we've gone through this year.

If you look at fourth quarter of this, give the expectations for a seasonally normal period, would you expect to get back to record levels for the units overall for the Company?

I haven't actually done that calculation, so I'm not sure I would give you the answer if I had it, but I haven't done it.

Because you're talking about a different kind of mix now.

So, the issue is how fast mobile grows and server grows versus the desktop share of the PC market shrinking.

And I haven't looked at it in that granularity, Mark, sorry.

And then just a follow-on then, Paul.

When you look at 2007 as a whole and you think about the manufacturing footprint that you have, what type of share gains are necessary to keep those fabs fully loaded and to prevent any meaningful increase in inventories from current levels?

That's a level of detail and share gain and unit projections that we're not comfortable giving out yet.

Then maybe just one for Andy if possible.

Sorry.

With the 45 nanometer costs starting to perk up here Andy, what should we think about the rate of ramp as we go through the first half of next year?

I won't give you anything new.

What I'll tell you is, we do expect it to be a little less than what we saw with 65 nanometer.

I'll give you more quantification in January when we start talking about margins for next year.

But I guess it's fair to say that it will -- from here we'll continue to increase in dollars on a sequential basis?

Certainly, through the first half of next year, yes.

And then I haven't looked at the third quarter yet but through the first half of next year, yes.

Okay, thanks, a lot guys.

And sir, we have a question from the line of Chris Caso with Friedman, Billings, Ramsey.

Thank you.

I wondered if you could talk about the incremental margins that you guys are seeing with the new microarchitecture products?

In other words, as those become a larger part of the mix, all else being equal, what do we think about gross margins on sort of a like for like basis?

Are they generally better margin products than the products that they are replacing?

Again, I don't want to forecast margin by product.

What we know is they cost less.

So, I'll tell you I expect to see costs trending down as those become bigger and bigger percentage.

As far as the pricing environment, again -- I don't -- we don't forecast ASP's for Q4, we're not forecasting ASP's for next year.

I can't give you that piece of it.

Costs will improve with the new products, yes.

And that's the case for each one of the products and each one of the categories, server, desktop, and mobile?

Pretty much.

And probably -- it may be the dye is a different percentage of the cost each of the pieces but the dyes are less expensive in all cases.

Okay.

That's -- and just as a follow-up.

With the head count reduction and the cost savings that come as a result of that, it looks like most of that is done by the end of this year.

As we look at '07, should we assume that most of that cost reduction is realized in the first quarter of the year and then some of it trickles out through the year?

So, I wouldn't make that assumption yet.

What we said is 95,000 people on board by the end of this year, 92 middle of next year, I haven't -- we haven't provided you a number for the end of the year, but my suspicion is it will be a little lower than 92.

So, you certainly see the biggest effects in Q4, Q1.

I think this program continues probably for the next four to five quarters.

Some of the ideas we have we are pursuing, some of processes we're reworking, we know will take anywhere from three to four quarters to get implemented.

So, yes it continues for a while.

Clearly the front end is the biggest end but I think you'll be able to see improvements, as I said, four to five quarters beyond.

Okay great.

Thank you.

And our next question is from the line of Michael Masdea.

Yes, thanks a lot.

In the enterprise market, and mobile and desktop specifically, you guys had a pretty good lock historically.

Can you talk, has there been any change in tha dynamic, any kinds of designs wins or customer wins from your competitor that are concerning you at this point?

Well, every one that they win is one less that we have.

So yes, any one would concern us.

But I think that we're very pleased with -- as I said earlier, the scope of the design wins on the new products both at large accounts, smaller new wave accounts and channel.

It's been, by any measure, the most successful and broadest product refresh we've ever had.

And having the capacity to take advantage of that is really the key to the whole new product structure working out for us.

And so, in other words, your current kind of view on market share suggests you're not going to lose any share and most of your CapEx is built upon that et cetera?

We are not giving you CapEx for next year yet.

But we're certainly not going to put in less capacity than we think we need to serve the market.

Got it.

And kind of follow-up on that point, when you think about your CapEx plan for next year, and I understand that you're not giving it to us yet, how much do you factor in your competitor's capacity.

Do you consider that or do you build it really just based upon what the market will bear?

Well, we certainly look at it, but as I said earlier, our view is that we factor in our lead on capacity.

So we enter the year so far ahead on advanced technology and we exit the year even farther ahead on advanced technology.

And I believe very strongly that you sell the best products first in this market.

And that's not -- that has not changed in 2.5 decades of the PC and it's not going to change going forward in my mind.

Got you.

And just to follow-up on the whole platform strategy, you've talked about in the past and more important going forward, you announced Viiv and vPro, etc.

Just give us kind of an update on where those stand and have they lived up to expectations?

I know some of them are brand new but --.

Well, Centrino is cranking away, we're in our third generation.

And we talked about what the fourth generation looks like at IDF by adding NAND and adding new chipsets and adding better a graphics and WiMAX.

Viiv is also in its second half of a year in term of shipments.

The product was significantly upgraded over the summer to enable remote wireless, remote broadcast of protected content to large screens around the house.

And there was a software upgrade for the installed base and new machines.

I have to say, though, that Viiv is probably suffering from the overall consumer market not being as robust as we first thought 12 months ago.

So, as our views of the entire market shifted over the course of this year, Viiv was one of the products that came down along with that.

Having said that, we're thinking that it's nicely poised for the holiday selling season and people are moving more and more of their content to these kinds of machines.

It's the best media machine out there.

vPro just started shipments and I'm very encouraged by the design wins.

But I think this is like any enterprise product, it has its own deployment cycle.

Great, thank you.

And our next question is from the line of David Wu.

Thank you.

The -- two quick ones.

First on the desktop, do you see a change in the competitive dynamics in Asia Pacific where your competitor is spending a lot of focus on that market and how quickly can -- would that market shift to CONRO, in your opinion?

And the other -- the last one is on the flash business, it's a [grab all] numbers.

But the operating losses are coming down despite lower revenue and I know that NAND flash is ramping over at Intel Micron and I assuming the losses are going to increase with time in '07.

How do you balance better profitability in the NOR and losses, presumably increased costs on the NAND side?

Let me take the first part and have Andy take the second part.

The APAC dynamic on desktops is one that, as I said earlier, really reflects the most heightened sense of competition in the marketplace today for a variety of reasons.

And CONRO, actually, had a earlier and stronger uptake in the channel than it did at multi-nationals.

Principally, because the channel sells more discrete based graphic machines than the multi-nationals do traditionally.

And we saw that take off nicely and it's one of the reasons we saw channel sales throughout the quarter, including the September month.

So, I'm encouraged with that product there.

And the channel players can move quickly, they can offer very good prices and they can try to be in the market before larger competitors and that give them their early to market advantage.

On the flash business, there are two separate, I guess, divisions inside the group.

You have the NOR and the NAND.

NOR, is quite frankly, making good progress.

Paul -- as we've noted, that revenue was down, that was a disappointment.

We lost some share.

But if you look at the other dynamics of the business, the spending is lower, getting more efficient, the cost per units are getting lower.

New products are coming out, which we think are topnotch products.

Get a little revenue growth there and that business is going to start to resemble a well rounded dynamic business for us.

So, I would round-up the NOR flash business, same goes.

Keep driving revenues, keep controlling costs, keep improving your manufacturing costs.

NAND is in its start-up phase.

We knew that when we approved the program.

I would say we're pretty much where we expected to be.

If anything, the loss may be a little higher, only because we think the opportunity is good and we're investing a little quicker.

For what you'll see in the P&L for a while, you're right, a semiconductor business that's in start-up mode, investing in engineering and manufacturing capability ahead of the revenue stream is a couple of pretty tight years.

Right now it's on our expectations and we're -- as a CFO I'd like to see better numbers quicker but this is what we expected and it's delivering.

And sir, our next question is from the line of Cody Acree with Stifel Nicolaus.

Thanks, guys.

Let me ask you about structural gross margins.

Andy, maybe for you with Dual Core, with more competitive AMD, if you look out of several quarters and really look at the kind of overall mix of things, how does the business look compared to maybe what you've seen in prior cycles when you've ramped into full production on new products?

So, I'm really not ready to give a long-term gross margin.

I'll probably never give a long-term one.

At the end of the year I'll certainly give more insight than I will now into the next few quarters.

What I'll -- I will talk about costs.

If you look at costs per microprocessors, cost per units and you look other than the time we were putting them on boards, little cards, the costs of the microprocessor doesn't change dramatically over time.

We did see one of the few spikes when we when with the first 90 nanometer Dual Core.

The new products, we're going to bring the costs back down nicely.

It's going to be pretty much in line with historical.

Possibly, again, a slight trend down.

So, where the costs are going, I'm fine with.

What will happen on the pricing side, which is the other half of margin has to do with, as Paul says, are our products the best, is their manufacturing technology the best?

Can we fill the high end of the stack?

How much capacity does AMD have?

What do they do with it?

So, those questions we need to wait and let some time pass before we answer.

The cost side, we can drive.

Great.

And then maybe follow-up on AMD's acquisition of ATI, can you talk a little bit about how that's changed your positioning -- not maybe positioning, but the way you look at chipsets and graphics going forward and how you're partnering?

And how is that changing the environment for you?

Well, it doesn't have a significant change.

Most of our processors go with our chipsets anyway.

And particularly coming out of last year where we had a rather acute shortage of third-party chipsets, we've made plans in terms of capacity planning to be able to handle a larger chunk of our own needs ourselves.

And I think you see our competitors similarly having a strategy that's trying to do the same thing.

The real issue is in the discrete graphics market.

Obviously, NVIDIA, I believe, will continue to support both parties.

I don't know what the plans are from ATI.

So from a discrete graphics standpoint, that may be the one element that we have to take a look at.

But I think there we're very comfortable with our ability to continue to grow the generational capabilities of integrated graphics by throwing Moore's Law of Transistors at it.

Any thoughts in taking maybe a shot at the discrete graphics again yourself?

Well, there's always thoughts.

If you're asking plans, I really can't comment.

Operator, we'll take two more questions.

Sir, our next question is from the line of Krishna Shankar with JMP Securities.

Yes.

As Core 2 Duo ramps up in the desktop space, do you anticipate that -- will that be one of the key factors in helping your desktop ASP stabilize going forward?

Or can you talk about the factors which could potentially lead to a desktop ASP stabilizing?

I think the CONRO ramp and the Core 2 Duo ram in general, are one of the reasons that Andy cited confidence on a firmer price environment.

And just as a follow-up, the Vista ramp, what do you feel about the adequacy of the Broadwater chipset to handle DX9 and potentially DX10 graphics and --?

Actually, Broadwater, is to my knowledge, is the first DX10 capable machine out there.

It is qualified for Vista Gold or it will be -- it is capable of handling Vista Gold when Vista Gold comes out or the high end of the line.

We worked very closely with Microsoft to make sure that happens.

The majority of computer users can have the best graphic experience.

Great.

Thank you.

And sir, our next question is from the line of Gus Richard with First Albany Capital.

Thanks so much for taking my question.

On the NOR flash business, the decision to stay in that business, is that more strategic?

Or is it P&L driven?

And if it's strategic, where does it fit in going forward?

That's a business that we believe has to provide return and pay for itself and that's what we're trying to get it reshaped in order to do.

As I've said, we're making progress.

We have more to go, obviously.

Got it.

And then in terms of improving performance, you've got 45 nanometers coming out next year and then a new architecture after that, which one could you sort of a handicap, which one is going to drive MIPS per watt, what's going to drive performance?

Is it going to more the processor or more the architecture?

It's both.

At IDF, I think I talked about, I know I talked about this.

We said that going forward from today forward over the rest of this decade we're going to deliver a 300% increase in performance per watt based upon that road map we put up there.

Some of that comes through the capability that the transistor budget gives you, which is one element of the technology transition.

And the other is we will continue to drive for lower power capability and lower leakage orientation at the transistor level, which helps the performance per watt independent of what the performance is.

So the new silicon technologies are being optimized much more than they were in the old days for mobility, for power friendly environments, for low leakage, that kind of stuff.

Got it.

Thank you.

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

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

Pacific Time tonight.

Those interested should dial 1-888-286-8010 and reference passcode 18401155.

Thank you.

Ladies and gentlemen, we thank you for your participation in today's conference.

This concludes your presentation and you may now disconnect.