超微半導體 (AMD) 2006 Q2 法說會逐字稿

Good afternoon, ladies and gentlemen.

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

At this time, I would like to welcome everyone to the AMD second quarter earnings conference call. [OPERATOR INSTRUCTIONS]

It is now with great pleasure to turn the floor over to your host, Mr. Mike Haase, Director, Finance and Investor Relations.

Sir, you may begin your conference.

Thank you and welcome to AMD's second quarter earnings conference call.

Our participants today are Hector Ruiz, our Chairman of the Board and CEO.

Dirk Meyer, our President and COO, Bob Rivet, our CFO, and Henri Richard, our Chief Sales and Marketing Officer.

This call is a live broadcast and will be replayed at AMD.com.

Telephone replay number is 877-519-4471.

Outside of the United States, the number is 973-341-3080.

The access code for both is 7600579.

The telephone replay will be available for the next ten days starting at approximately 5:00 p.m.

Pacific time tonight.

In addition, I would like to call to your attention that our third quarter 2006 earnings quiet period will begin at the close of business Friday.

September 15th.

I would also like to remind everyone that our next analyst day is scheduled for the morning of September 14th in New York City.

We'll provide you with more information closer to the event.

Before we begin today's call, I would like to caution everyone that we'll be making forward-looking statements about management's expectations.

Investors are cautioned that our forward-looking statements involve risks and uncertainties that could cause actual results to differ materially from our current expectations as set forth in the forward-looking statements.

The semiconductor industry is generally volatile, and market conditions are particularly difficult to forecast.

Because our actual results may differ materially from our plans and expectations today, I encourage you to review our filings with the SEC where we discuss in detail our business and the risk factors that could affect us.

You'll find detailed discussions in our most recent SEC filings, AMD's quarterly report on form 10-Q for the quarter ended March 26, 2006.

With that, I'll turn the call over to Dirk Meyer.

Thank you, Mike.

In spite of our continued strong year-over-year growth, the second quarter leaves us unsatisfied as we did not meet our sales objective.

On the positive side, we achieved 53% year-over-year sales growth, 57% gross margins and recorded the 12th consecutive quarter in which the year-over-year microprocessor sales growth exceeded 20%.

Even in this challenging market environment, we saw significant progress towards our goal to expand and strengthen our relationships with our largest global customers.

Highlights of that progress include: Dell announced their intention to offer dual core AMD Opteron based servers by the end of the year, Lenovo announced plans to offer AMD powered Think Center commercial desktops, Fujitsu Siemens computers launched two new AMD based commercial offerings, Sun Microsystems launched AMD powered commercial claims, and we continue to enjoy strong commercial growth across all product categories with Hewlett-Packard.

Our strong gains with our biggest global customers were due in large part to the continued adoption of AMD Opteron processors.

Once again, we set a record for AMD Opteron revenues with increases of 26% sequentially and 141% year-over-year.

ASP has increased and we believe we continued to gain unit and dollar share as the server market continues to embrace AMD solutions across a broader industry footprint.

We made further progress in manufacturing, allowing us to efficiently balance customer demand with the unique dynamics of our industry.

More specifically, Fab 30 operations continued to perform at industry-leading levels.

Fab 36 is ramping ahead of plans.

We commenced shipment of products out of chartered semiconductor and we're on plan to ramp 65 nanometer production in the second half of this year and we expect to do so at mature yields.

Furthermore, we have announced plans to further expand production capacity at Fab 36, convert Fab 30 from a 200 millimeter to a 300 millimeter tool set and accelerate our processor technology road map with an aggressive transition to 45 nanometer within 18 months of 65 nanometer production.

AMD also received a nonbinding $900 million cash incentive package from the state of New York to build a wafer fabrication facility in Luther Forest, New York.

On behalf of AMD, our customers and partners, I want to thank the state of New York for its tremendous commitment.

In early June, many of you attended our technology analyst day where we detailed our Torrenza strategic initiative.

Torrenza represents a significant shift for our industry.

It is industry's first open X86 innovation platform, enabling processor and hardware vendors to innovate within a common ecosystem.

For our customers and partners, Torrenza will create an opportunity for differentiation and value that's unprecedented in our industry.

Leveraging the scalability of the direct connect architecture, we also plan to demonstrate our next generation microprocessor core in a native quad core implementation before the end of the year.

In summary, while we cope with short-term disruptions in the market, we did not and will not lose our focus on executing our long-term strategy.

Our global customer base is growing, supported by our industry leading manufacturing execution and truly, game changing technology.

Now, I would like to ask Bob to review the results of the quarter as well as to provide the outlook.

Thanks, Dirk.

As we stated in our press release and as Dirk just commented, we're disappointed that we didn't hit our top line guidance.

However, we were very pleased with several elements of the quarter.

As a result of the Spansion IBO, by comments comparing the second quarter of 2006 to the second quarter of 2005, do not include the results from what was our memory product segment.

Second quarter sales of 1.22 billion were down 9% compared to the first quarter of 2006.

Sales were up 53% compared to the second quarter of 2005 driven by growth in all segments.

Geographically, processor sales were sequentially higher in China, offset by lower sales in North America and Europe.

Consistent with normal quarterly seasonality, our total microprocessor unit shipments were down 4% sequentially, driven by :lower shipment of client units and lower client ASPs.

High volume desktop processor ASPs were challenged in the quarter due to deep discounting in the marketplace and in some cases we walked away from that business that didn't make sense.

We were pleased by the growth in our server processor business, recording sequential double digit percentage unit growth and single digit percentage ASP growth.

We believe we once again took market share in the server space with 141% year-over-year AMD Opteron sales growth and sequential growth of 26%.

We're well on our way to achieving our year end goal of more than 30% server market share.

Gross margin was a solid 56.8% and was within our long-term guidance.

Down slightly from the record of 58.5% in the first quarter of 2006.

This was largely due to lower desktop processor ASPs.

Total operating expenses, which include R&D and SG&A, were up 13% from the prior quarter primarily due to the extra week of operations in the quarter and increased marketing expenses in support of our long-term goals to acquire new customers, expand business with existing customers, and increase commercial sales.

Operating income of $102 million was down from the prior quarter, an increase in the second quarter of 2005.

Net income for the quarter was 89 million or $0.18 per share.

These results included: noncash employee stock-based compensation expense of $18 million or $0.04 a share, a $12.5 million share loss or $0.02 per share associated with our 38% ownership in Spansion LLC, and a net gain of $10 million or $0.02 per share as associated with Spansion LLC repurchase of its 12.75 senior subordinated notes.

Now, let me turn to the balance sheet.

Our cash balance was $2.5 billion in the quarter, down $100 million from the first quarter due largely to increased capital expenditures associated with investments in Fab 36.

Inventories increased 68 million as planned from the first quarter driven by the outstanding ramp of capacity of maturity of Fab 36.

Days of inventory ended the quarter at 76 days.

Now, let's talk about the outlook.

AMD's outlook statement for the balance of the year is based on current expectations.

The following statements are forward-looking and actual results could differ materially depending on market conditions.

AMD expects demand for its products to be seasonally strong in the second half of 2006 and we expect third quarter sales to increase sequentially.

As a reminder, we will return to our typical 13-week quarter for third quarter.

Operating expenses for the third quarter are expected to be flat from the second quarter.

Noncash stock-based compensation expense in the third quarter is expect to be approximately $20 million at the current share price.

Depreciation and amortization will be approximately $200 million in the third quarter and $800 million for the year down slightly from our guidance of 2006 of $825 million.

We expect 2006 capital expenditures will remain at our prior guidance of approximately $1.7 billion. 2007 capital expenditures expected to be approximately $2.5 billion as we expand Fab 36 capacity capabilities and start investing in our conversion of 300 millimeters of Fab 38.

So, in summary, we are confident about the second half of the year.

And we continue to focus on the successful execution of our strategies.

Now, I'll turn it over to Hector.

Thank you, Bob.

In this past quarter, we have entered into a new stage of industry transition.

The world has spoken.

And it's really a new world that's about differentiation.

It is a world about choice, about fair and open access to market.

It is not about benchmarks.

However, it is about innovation.

Innovation beyond the transistor and beyond the core.

Innovation based on customers and their needs and I believe this transformation is irreversible and the company with its strongest leadership fundamentals will emerge as a partner of choice.

We continue to expand our platform footprint with the world's most discerning customers.

Top tier global OEMs, like Acer, HP, IBM, Lenovo, Sun and soon, Dell.

With truly innovative offerings in site, like our Torrenza open innovation platform, our upcoming four by four enthusiast solution, and the industry's only native quad core, we will continue to provide a new path to differentiation and sustained value creation for companies starved of this choice.

We are scaling our capacity to meet the growing number of companies wanting to work with a more customer-centric partner.

Our Dresden and state of New York announcements secure our ability to meet significant increases in demand for our world-class products.

Our commitment to Internet enable 50% of the world's population by 2015 took another step forward this quarter with our agreement with Microsoft to support flexible business models in the emerging markets.

Our industry has recognized that one business model, one technology, does not fit all.

We continue to be a magnet for the world's best and brightest technical minds and it is because of our exceptional collection of talent that I'm confident in our ability to no longer just survive these industry disruptions but thrive in them.

Our industry is one of the most important in the world and as such, we have a responsibility to evolve and transform along with the needs of our partners, customers, and markets.

This quarter, we witnessed a few aftershocks from this transformation as our new approach to innovative technology and serving customers continue to take hold.

And we're fully confident and look forward to being able to thrive in this environment and continue our momentum to gain share.

Finally, I would like to thank AMD employees everywhere for their commitment and excellence, even in a challenging quarter, they have remained focus on our most important objective, delivering customer-centric innovation that enables our customers and partners to win in their markets on their terms.

Now, let me turn it over to Mike Haase to start the question-and-answer period.

Thanks, Hector.

Operator, if we could start the polling, please.

Thank you. [OPERATOR INSTRUCTIONS] Your first question is coming from Mr. Tim Luke with Lehman Brothers.

Please go ahead.

Thank you so much.

I was wondering if you could give us some framework for some of the different puts and takes we should think about with respect to the gross margins as we move into the third quarter.

Sure.

Obviously we forecasted -- we think sales will go up so you have to play that through the equation.

Fab 36 will continue to ramp but all of the inventory will be valued.

It is ramping flawlessly as I kind of said.

So, gross margin, we don't give a forecast on gross margin but to me, everything is moving in the right direction.

And with respect to the sales being up, if you could just remind us of how you perceive seasonality and whether, within that, we should expect this mix of maybe lower ASPs balanced by strong years and everything other than Opteron or whether after such a strong ASP in Opteron in the second quarter, we might expect that to moderate some in the going forward.

Thanks.

Tim, this is Henri.

I think it is clear the environment will probably remain challenging in the desktop.

However, I'm confident that our brand positioning and the positive impact of the dual [core turion] launch will help us in the mobile space, and as you pointed out, we should expect some offset with a strong continued success of Opteron.

So, the desktop and mobile are up with both having ASPs lower and possibly --

No, no, I would say units in both mobile and server ups with solid ASP and frankly, the market is uncertain on the desktop as we've seen in the second quarter.

I won't make a projection there.

Is it fair to say you've seen your large competitor as aggressive in the early part of this quarter as they were at the end of last quarter?

You know the issue that we're seeing in the marketplace is not so much the level of aggression but the way it is being communicated and how it is disruptive to the partners.

Thank you very much, guys.

Your next question is coming from Mr. Adam Parker, of Sanford and Son.

Please go ahead.

Sanford and Son, I think that's a TV show.

Anyway [laughter].

I know you don't want to talk about benchmarking, Hector but can you just talk to us about -- Henri, can you talk about your technology in the desktop business a little bit?

How do you think your products here stack up against Intel in the desktop space as we kind of head into comparing ourselves to Conroe and what can you do to gain share in desktop here in the next six months?

Well, it is very clear that our customers have a much better knowledge than I do of both the existing and future products that AMD and our competitor are offering.

All I can tell you is that I continue to see a strong trend of increasing the number of AMD platforms they will offer to the market.

That tells me that more than ever, we're the smarter choice.

Is that predicated on superior technology or pricing or both or to what do you attribute that?

It is Dirk here.

Maybe I can provide some insight.

It's interesting, from my perspective, the performance crown in the microprocessor business is something that's been passed back and forth several times over the past ten years.

I expect that to continue.

Interestingly, who has the performance crown at a given instant is relative---is relevant really to a pretty small fraction of the marketplace.

That fraction is one we refer to as the enthusiast market.

Interestingly, our four by four platform initiative, we think is going continue to hold the hearts, minds and wallets of the enthusiast in the industry.

But more relevant, really to our business is the so-called mainstream market where we have always and will continue to provide technology solution that meets the needs of that marketplace and we do so frankly without the monopoly [attacks] of our competitor.

Our customers understand that.

Their customers understand that.

They all want choice.

And they're making that vote.

They have made that vote and we think they'll continue to make that vote.

Ok.

Lastly, guys, can you talk about what -- how this extra week impacted your revenue in Q2 and how you expect it to impact your revenue in Q3?

What's embedded in your guidance here about the one less week on the sequential basis in Q3, please?

Well, clearly -- this is Bob -- clearly, we didn't get any uplift for the extra week in the quarter.

In the second quarter from that standpoint.

Clearly, we have the cost side of the equation.

For the depreciation payroll, et cetera, et cetera, associated with the extra week.

Third quarter goes back to the normal 13 weeks and we expect to grow from second to third, period.

Cost will come down which is why we're forecasting flat, a quarter on quarter because we lose that extra week.

Adam, this is Hector.

Let me just add a couple of points there.

One is as Bob pointed out, I want to highlight it, we expect our revenue to be up in the third quarter.

And despite the 13 week, one less week of work.

But in addition to that, as we said in our press release, we expect the second half of the year to be seasonably strong for us and the reason we feel that way, I think it is important to know, this is not just an analysis at a pretty high level.

This is based on the inputs from our customers as Henri and Dirk pointed out that are telling us that the [indiscernible] availability of product, et cetera.

They feel very bullish about their expectations of acquiring products from us.

We feel very positive about the second half of the year.

As a matter of fact, I believe that in the third quarter, we will be able to be above seasonality.

The reason I was asking you is because you talked about share gain and server and then you didn't mention what you think happened to share in both desktop and notebook in the second quarter, and I just--- If Intel is guiding to 7.5% growth in Q3, you guys are guiding to growth but you have one less week in Q3, I was trying to figure out if you were taking out the extra week in Q3.

Do you think you'll grow in-line with the market or faster and continue to be a share gainer on the client side in Q3 and --

it is clear---- probably two or three of us want to make a comment here.

It is pretty clear that we feel very bullish on the server side.

It's pretty clear the numbers are big. 26% sequential growth and 141%, we are gaining share.

In the desktop and notebook --

we'll continue to do that.

Now, let's see what the third part is, numbers come out in a few weeks relative to the client because frankly, Adam, if you look at the numbers right now, it would be very hard for us not to conclude we might have held or perhaps even gained share on the client, despite the rather challenging environment that we saw in the second quarter.

I want to remind you of what we all have said is that our plan are to always gain share across the board.

That has been our goal -- our goal is to break the monopoly and we continue to have that goal.

We've had 12 consecutive quarters of growing more than 20% year in year for 12 consecutive quarters.

Our plans are to continue to do that.

So, I feel good.

Thanks for your time, guys.

Thank you, Adam Parker from Sanford Bernstein.

Your next question is coming from Mr. Mark Edelstone with Morgan Stanley.

Please go ahead.

Good afternoon, guys.

Couple of questions, if I could.

The first one, I guess for Bob or Hector, what does the mix look like in Q3 between 300 millimeter wafers and 200 and then what does that look like in Q4?

Hi, Mark, it is Dirk here.

Rather than give you the specific mix, why don't I give you the overall ramp of Fab 36 and you can deduce from that?

We started the ramp roughly at the end of '05.

We'll complete the ramp to the full 20K per month level roughly end '07 and it is a pretty linear ramp across those eight quarters.

Ok.

Fair enough.

And then maybe for Henri or again for Bob, perhaps, I think your Op Ex growth in Q2 ended up being more than you had expected.

The guidance was eight if I remember correctly.

Can you talk about some of the programs you put in place and what, if anything, was a surprise to drive the overall spending levels above what you had initially anticipated.

Yes, I'll give you just a high level and I'll turn it over to Henri and he can get into a little more specifics, but basically, we did a fair amount of investing in the future, particularly in the marketing area to continue to power our way through this environment with our eyes still set on the market share gains that we've laid out for us in all of the different product categories.

But, to give you a little more color and specifics, I'll turn it over to Henri.

Thank you, Bob.

You know, Mark, we've always had an aggressive plan to penetrate the commercial segment in 2006.

We made no secret of that.

Fundamentally, based on customer feedback, the increasing number of platforms that will be available on the revision Our Opteron server platform which is coming out next month and the opportunity in front of us in the second half of the year to continue to accelerate our momentum in the enterprise, I made the decision to also accelerate our spending.

And I'm sure you've seen our power campaign and some of the other campaigns we've put in place.

We felt it was the right time to invest in the success of our enterprise penetration and the bulk of our spending is in that area.

Two very brief ones if I could.

One is what do you expect ASPs to do in the third quarter and Hector, you talked about seeing seasonal strength in the second half of the year.

How much of that strength do you think is coming from normal seasonality versus share gain?

Mark, we believe that because the environment that we called challenging in the second quarter is going to continue in the third and probably likely even into the fourth, we're going to be facing the second half where ASP projections are going to be challenging to make due to, frankly, the uncertainty of what our competitor is going to do in the desktop space where a lot of the challenge has been.

So, [indiscernible], the ASPs were going to be reasonably flat, ok.

As far as our reasons for optimism relative to growth is frankly because our customers are telling us that they expect the demand for our product so this is not a statement of the market, this is a demand for our product, is healthy.

And that that combined with what we believe are some customer acquisitions strategies for the second half of the year because we still have customers around the world that are not yet purchasing AMD products, the combination of that is -- makes us feel optimistic, i.e. translate that to be is we're going to gain share.

Thanks a lot, guys.

Thank you.

Your next question is coming from Mr. Eric Gomberg with Thomas Wiesel.

Please go ahead.

Did you say that you actually expect to be above typical seasonality in the third quarter and if so, what is typical seasonality?

No.

We don't know what seasonality this quarter is going to be.

Frankly, it changes every year.

It ranges from lord knows what number.

I forget.

Average of probably in the --

--- typical is four to seven.

If whatever the seasonality is this year which is a function of so many things, if this year happens to be a third quarter that's seasonably up X percent, we expect to do better than that.

Ok.

Fair enough.

Could you talk about where you expect inventory on your balance sheet to be exiting the third quarter and also what you think the status of channel inventory is right now?

Yeah, this is Bob.

Our plan is to build inventory in the third quarter in anticipation of an even stronger fourth quarter.

Probably at a slightly lesser rate than we did in the second.

Fab 36 will be where that build will take place.

We've been running at full capacity in Fab 30.

So, fourth quarter will drain that off.

Typical pattern is build in second, build a little bit in third and drain off inventory in the fourth.

We're pretty consistent on that pattern and that's where we're going.

This year might change in the fourth.

Depending on how the customer input comes in for the start of the '07 period of time.

As far as the channels today?

Henri, you want to make comment?

Sure.

I'll be happy to.

The channel is obviously where most of the unrest has taken place in the second quarter.

So, consistent with our general practice, we've kept our inventories as low as possible and then build inventory during the quarter in the channel.

And given Intel's behavior in the channel, would you expect to continue to walk away from low end business this quarter or do you expect that to moderate?

Well, Hector has always been very clear our goal is to make profitable growth.

So, we're going to take business that makes sense for our customers and for us and we're not going to go chase, what I call lighting a cigarette in front of a gas leak.

Ok.

Fair enough.

Thank you.

Thank you.

Your next question is coming from Mr. Joseph Osha with Merrill Lynch.

Please, go ahead.

Lighting a cigarette in front of a gas leak, huh?

So, I have a sense as to how Fab 36 is going to ramp.

You said 20K a month by the end of '07, right?

But as I get beyond that and I think about the Fab 30 coming down and then coming back up, can you give me a sense of say the end of '07 run rate, what's going to happen?

Will 30 be coming down during '07 and coming back in '08?

I'm trying to get my arms around the total output.

At the high level, what happens is we ramp Fab 36 to 20K and then actually beyond that in '08, ok.

Ok.

In '07, we start to take down Fab 30 and convert to a 300 millimeter tool set.

Ok.

And in addition, we've got chartered semiconductor which is ramping now.

Is Fab 30 off-line by say the end of '07?

It never really goes off-line.

It is a conversion.

Ok.

But I mean it is a whole different tool set.

I would think it would affect --

Think of it this way, Joe.

We'll probably never go below 50% utilization in the facility as we continue to flip out tools, what we're actually doing is building a separate building to actually augment the capacity to be able to flip the tools through the system so we'll never actually go below about 50% utilization of the facility in the worst quarter of time.

But again, we also have charter to fill in the blank from that period of time.

We see continued capacity expansion through the '07-'08 period of time, not exactly linear but darn close.

Ok.

So, let's just imagine it is the end of '08 and I've got 36 running at let's call it 25,000 wafers a month, maybe the retrofit of Fab 30 at, call it 20, and could charter be--I mean is charter ever going to be that big or would charter be maybe ten or something like that?

It will be modulated.

Ok.

But you could be in enterprise with mid '08, 50,000, 300 millimeter wafers a month.

Yes, sir.

Thanks a lot.

Your next question is coming from Krishna Shankar with JMP securities.

Please go ahead.

Yes.

Can you comment about the pricing environment here, so far in July.

Do you feel that some of this was based on your competitor blowing out [indiscernible[ or getting obsolete and will we see a more rational environment in the channel going into Q3 and Q4?

Well, I think there were several factors but probably the biggest factor was the fact that it seems the competition decided to throw away one of their brands and completely reposition pension.

That had a very negative contribution to the environment particularly in the channel.

It is a little early now in July to give you a forecast of what's going on.

Our checks indicated there is still a lot of inventory out there and that's why, in my previous comments, I was fairly optimistic about the mobile space and the server space.

I'm cautious about the desktop space because that's where the bulk of the channel business exists.

My follow-up question is I think everybody saw the reviews on the competitor's new product line and do you folks have any comments on what you've seen in terms of the trad reviews and what your upcoming, a socket two past, A and two socket will do in terms of the advancement of performance and power?

Of course we're looking at all of those reviews.

As you know, the main feature that the AM2 socket is bringing to the market is the support of DDR2 memory.

We've never presented it as a performance improvement.

We are going to continue to provide faster products to our partners.

I'm not going to disclose at this point in time our Q3 and Q4 plan.

We're looking at what the competition are doing.

They seem to be innovating at the core.

We're going to continue to innovate both at the core and the platform level.

I would just like to add that when we launched Turion, the first launch was very successful, if you remember, back to the days we actually stated that that was with the most successful product launch we had and fastest ramp.

We're now -- when we introduce the next two, it is doing better than that and the reaction from our customer base is very, very strong toward our product.

My final question is when will Dell actually start to ship the Opteron server base products in volume?

You should ask them.

What they publicly stated is they would -- their plans are to have a product by the end of the year.

On the server.

Does that mean that there's other products?

You seem to be hinting at a broadening of a relationship with Dell.

Our hopes are always to broaden our relationship with all of our customers.

We used to be mainly a desktop supplier to HP.

We're now a broad-based supplier to HP.

That is our plan, our intent.

Our desire.

It is no different with Dell.

At this point in time, the only product they have openly announced is the launch of a server by the end of the year.

Thank you.

Thank you.

Your next question is coming from Lacie Higgins with Prudential.

Please go ahead.

Hi, actually, it's Mark Lipacis from Prudential.

A couple of questions, Dirk, I thought in the beginning in your comments, you talked something about a native quad core and I missed that.

Would you be so kind as to just review what you were talking about and is that a change in -- of the production schedule for quad core?

High level is not really a change.

It is really an amplification to what we said at the analyst conference in June.

We've got a new core under development.

The first in stantiation of which will be in a quad core form.

To be launched roughly mid '07 and what I said is we'll demonstrate that by the end of the year.

Ok.

Thank you.

And the -- Bob, for the $2.5 billion in Cap Ex for '07, what is the expectation for funding that through cash flow from operations or through some other means.

No, current cash on the balance sheet and cash flow from operations.

Ok, great.

Last question, as you guys are ramping Fab 36 and could you give us a framework for understanding how we should think about the cost structure on products that are coming out of Fab 36 versus Fab 30 right now.

Thank you.

Sure.

Right now, we're still Fab 36 is a premium, meaning costing more than Fab 30 because we're still to Dirk's point, we're still quite a ways away from the appropriate asset utilization in the space and kind of to decode what Dirk said, we'll go out of this year about 50% utilized from a wafer start standpoint.

The 300 millimeters, as we all know, yields about a 30% advantage compared to a 200 millimeter wafer and we'll start crossing to Fab 36s cheaper at the same technology node in the fourth quarter kind of time frame.

Great.

Thank you.

Your next question is coming from Mr. Michael McConnell with Pacific Crest.

Please go ahead.

Thank you.

Looking at your competitor's balance sheet yesterday, we saw inventory up 22% sequentially.

Now at record inventories.

Your inventory on your balance sheet is up 20%.

You're both giving indications you're going to raise inventory on absolute dollars again in Q3.

Both saying you're going to be gaining share.

Why shouldn't we be worried about that this could get even uglier as we progress in the second half of the year?

First of all, let me -- a couple of comments.

I believe that the situations are quite different.

It is clear from the actions taken in the second quarter that there is an awful lot of aging inventory that needs to be dealt with and we recognize that's going to be a challenge and creates this very challenging environment that we talked about in the second quarter that we think will continue in the third quarter.

We are not building aging inventory.

We're building inventory like Turion X2 which is highly [indiscernible] by our customers.

Our up to demand is strong.

We've shipped everything we built in the second quarter in Opteron.

So, our view is somewhat different and even if our competitor is building inventory on the new products that they have announced, they still have the very large aging inventory that you have to deal with.

So, I believe these are very two different things.

Similarly, when we look at our capacity plans and expansion, they're based again on the inputs our customers give us on their wishes and desires to participate in the products that we build.

And I said in my opening remarks, which we believe very strongly if a fee change in the industry is that customers really have embraced the idea of choice.

And they want and need AMD to be a major player and a major supplier.

I believe our performance over the last few years has earned us the right to be a major supplier to its customers.

We believe that's going to increase in the footprint with each of them, it's also going to increase.

So, I believe our plans are sound.

Our expansion capacity is very sound and based on a planning based on customer needs and from our end, we are not worried about having too much capacity.

Fair enough.

And then looking at -- in September, there are some indications that your competitor is going to be changing their pricing strategy, at least traditionally, with the rebate activity with the OEMs versus the channel.

Now, understanding you're taking some share at the OEMs currently but looking at the channel, that's historically been a very strong segment for you.

With them now looking at potentially evening out the pricing between the channel and the OEMs, are you still comfortable with that change that you can still take share in the third quarter as well as the fourth quarter?

The information that you're referring to is supposedly being implemented by the competition on the 24th of July price move.

But all our indication in the market are that this is -- this is a story or smoke screen and that there's no fundamental change in the price structure of OEMs versus distribution.

Thank you very much.

Your next question is coming from Michael Masdea with Credit Suisse.

Please go ahead.

Thanks a lot.

I guess, this is the one question follow-up about competitive environment.

It sounds like the real battleground on the desktop area and the whole idea of going after profitable growth.

Is there some tipping point at which you say the share loss is too much and you're willing to play that game more or with Fab 36 ramping up, can you get more aggressive or is that always going to be the case where you'll walk away from too aggressive desktop business?

It is Dirk here.

I'll start and I think Hector is going to want to add.

First of all, our high level approach is to build capacity based on the demand signals we get from our customers.

Within a capacity context, make decisions and maximize the gross margin dollars we get for the business.

That's what we'll continue to do.

Relative to the pricing dynamic in the marketplace; from my perspective it's kind of interesting, we have always had price competition across the entire front of our business.

I think what's changing is in our -- Intel has not always had price competition across the entire front of its business but is beginning to feel that change.

So, frankly, the idea of continued price competition doesn't bother us because it is nothing new to us.

The only thing I would add is a minor modification which is important.

You said the desktop will be the battleground and I think I will add to that is right now, this challenge in environment is desktop in the channel.

And that -- as Dirk pointed out, we welcome competition and we thrive on figuring out ways in which we can thrive in that environment.

The one thing I do want to point out that Dirk and Bob alluded to in earlier comments.

They're so important I want to make sure they don't get missed is we managed to achieve in the first and second quarter, a gross margin performance that's between 55 and 60% which is quite strong and frankly, levels we hadn't achieved in our Company for a long time.

And we have managed to do that being at 200 millimeters, 90 nanometer technology and a factory that was small in the relative terms.

We're now moving toward -- and by the way, that was prior to us participating strongly in the server space.

Therefore, our ASPs were even lower.

So, we're now moving in this second half of the year.

Aggressively toward 300 millimeter, aggressively ramping 65 millimeter and because of our stronger participation in both the server space and the commercial space, improving ASP outlook once we get past this nightmare of a challenge in environment.

So, everything we're doing now is in the direction of goodness based on an already strong performance based on the parameters I just mentioned.

So, we believe that our opportunity to continue to be very competitive, even in this very challenging desktop environment is only going to improve and get better.

Great.

Thank you.

And then kind of in that vein, you had some pretty significant wins in the commercial market on the desktop side of Lenovo and [indiscernible] Fujitsu.

Walk us through how you kind of got over that hurdle to get into the commercial markets and how do you get the OEMs and some of the ones you haven't gotten so far, to expand their commercial offering so you can have more access to those higher ASPs?

Well, we've always told you that for us to challenge is not the demand at the end user level but the fact that the market would be offered the choice.

And it is very clear to me that it is increasingly obvious to companies in our industry that offering choice is a competitive advantage for them.

Their sales force talking to their customers.

And Hector talked about a change in the industry and that's what -- there is a tipping point frankly, the penetration of Opteron and the enterprise base has helped us give those customers confidence that there was true demand for AMD products.

Now, we have to continue to ensure that we penetrate all of the OEMs and offer a greater number of sku's and that will drive a greater market share.

It is all about having access to the market and breaking the monopoly.

Great.

Thanks a lot.

Your next question is coming from Mr. Jim Covello with Goldman Sachs.

Please go ahead.

Good afternoon.

Thanks so much.

Quick question on the Cap Ex.

I understand that today you don't believe that you have or are adding too much capacity.

I understand your '06 and '07 Cap Ex budgets are a sign of your confidence in the business but there comes a point at which if you were to miss numbers a couple of quarters again in a row, the capacity and the Cap Ex ramp would become a problem.

The question is at what point would you think about scaling back on the Cap Ex understanding that today you don't believe there is a problem.

Under what scenario would you think about scaling it back?

Well, I'm not sure I would lay out at this point a scenario.

But I -- hopefully we've demonstrated historically to date, we're pretty flexible in this area.

It is not just a bill it yourself -- build it and damn the torpedo kind of issue.

All of those numbers are relatively flexible--- to flex them up or down depending on the needs of the marketplace depending on the signals we're getting from our customers.

Do you think there is a scenario that the Cap Ex could be adjusted within 2000 -- the 2007 Cap Ex plans could be adjusted within 2006 or do you think it would need to be a longer time frame over which you would need to see weakness relative to your plan in order to adjust the plan Cap Ex?

We think we can affect future periods at any moment in time by pushing and pulling on those numbers.

So, we could be in '07 and affect '07 numbers just like we're in '06 affecting '06 numbers because we pulled it in and actually have accelerated our spend to put more capacity in place.

So, we're pretty flexible in that term and then in addition to that, we've got the toggle of charter semiconductor to flex up or flex down.

Right.

Then final question but same topic, how much near term pain are you willing to withstand in terms of Cap Ex or depreciation increasing in a weaker revenue environment?

How much of a near-term decision is it versus a two or three quarter decision?

In other words, if you started losing money, would that be alone enough to cut -- make you cut your Cap Ex plans for the future?

Or would even then, it would be dependent upon what your future outlook would be?

Well, frankly we hate hypothetical questions and I think because we don't know what the environment is going to be when one of those parameters gets met.

I think we're prudent in what we do.

We've demonstrated that in the past and as Henri pointed out our commitment is to profitable share of growth.

We're careful with that.

Having said that, we're not going to let a quarter or two of challenge and environment detract us from our long-term vision of breaking the monopoly.

So, we'll put in place and shoot ahead of the duck and the only thing that can keep the duck from getting there is the monopoly.

Thank you very much.

Thank you.

Your next question is coming from Mr. Glen Yeung with Citigroup.

Please go ahead.

There was a question earlier that asked whether or not there was some share level at which point you would capitulate.

Maybe ask the question another way.

Is there some margin level at which point you say you know what, I don't want to go below that margin level and that's all there is to it?

Again, these are hypothetical questions that have way too many parameters.

First of all, we have strategic customers, we have strategic partners.

We have regions of the world where we have an in turn strategic intent to do certain things.

Those are rather complex mixture of challenges.

In relation to that, we have brands we have created that address different segments, different markets.

It is not a simple answer.

I think what I would like to leave you with is we have a competent executive team here that manages the cause, manages the brand, manages the relationship with customers.

I think we do it well and frankly, I think that's one of the strengths going forward is that that's just not going to change.

Good point.

Next question I have is -- this may sound a bit odd given the environment that we're in but I'll ask anyways which is do you actually sense that given -- where you are today, as compared to where you were a quarter ago, and recognizing that there wasn't a lot of things that you necessarily knew were going to happen in the next 90 days a quarter ago, do you feel like your visibility now is a little bit better?

You live with a quarter of Intel's craziness.

You've got some interesting share gain opportunities coming up for you this quarter.

Seasonably it looks better.

Does it feel a little more certain to you or maybe just your sense on that?

So, first, let's remember that the second quarter in our industry is always the toughest.

So, for me, to give you visibility to the third quarter and the fourth quarter where you have back to school, you talk about the consumer market.

Back to school in the third quarter.

Christmas season in the fourth quarter, things become a little more predictable because those are stronger quarter inherently assuming industry seasonality would be the usual one.

Secondly, there is only so many brands that a competitor can throw away.

Our competition destroyed in less than six months, two of the most strong brand they've ever built, Intel inside and pentium.

I don't know how many brands they're going to throw away but they'll run out of them.

Therefore, as a result of that, there is a little bit clearer picture in terms of what could be the strategy.

Frankly, the issue that we're dealing with is that all of this disruption is not expending the market.

There is not a lot of elasticity in some of the pricing action that has been taken as demonstrated this last quarter in the channel.

And the partners are a little bit confused.

I think it is time to settle but settle down and give them long-term guidance.

That's what we're doing.

We don't want our partners to be worried about what they're doing, putting our products in the marketplace.

And we want to be the predictable, reliable partner and we're going to be that in the second half of the year.

That's actually a very helpful answer.

Thanks.

I just have one last question which may be a little esoteric here.

I want to get an update on your ZRam technology and if there's been any progress there.

Yes, it is Dirk here.

We're continuing to evaluate that technology internally and aren't prepared to comment further.

Ok.

Thanks a lot.

Your next question is coming from Mr. David Wong with A. G. Edwards.

Please go ahead.

Thank you very much. 65 nanometer roll-out, can you give us an update on when you expect that to be?

We'll start ramping the factory in the second half of this year and ship into the market our first parts by the end of the year.

Great.

Charter--you've mentioned charter a couple of times.

Can you give us some feel for the growth margin and profitability for stuff that you ship from charter?

I don't think we want to be specific as to profitability and gross margin numbers, the commentary I'll provide are we're shipping and planning on continuing to ship a mix of client products from chartered semiconductor.

The only thing I would add, I think it is appropriate to say that we are very, very happy with charter and their management team.

They've done an outstanding job and the yield of the quality of the product is very good.

Could we assume the gross margin would, by definition, be lower than in house?

No.

That's a true comment.

They need a taste of the action to be in the game.

We have no capital investment in the charter semiconductor business, so we have to pay for the cost of capital and the wafer but it is all marginally good.

It is incremental margin dollars to the Company to run the charter semiconductor business.

If it wasn't, we wouldn't do it.

Yes.

That's great.

My final question, I was interested in your comment that you've got marching orders not to take any unprofitable business and also about shooting at the duck.

If you had a big particularly important strategic duck that was quacking at you, would you take unprofitable business? [laughter]

No.

The marching orders are, we have to be prudent in the things that we did and we had to have profitable growth.

Sometimes addressing a unique particular need with a customer is very strategic could be a very narrow segment that may not be profitable in totality that customer segment or market has to be profitable.

Great.

Thanks very much.

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

Thank you.

Your next question is coming from Mr. Chris Danely with J.P. Morgan.

Please go ahead.

This is Sean Webster for Chris.

Had a question on the operating expenses.

Can you expand a little bit on why they might be flat sequentially compared to a 14-week quarter and then I have a quick follow-up.

Sure.

We lose a whole week.

I mean so a whole week, pick your number but in round terms, that's a 7% type issue so there is a 7% of cost that goes away when you transition from a 14 week quarter to a 13 week quarter.

So, basically, we're saying we'll have a 7% increase in natural cost offset by the 7% decrease of the extra week.

Ok.

Can you maybe talk -- characterize it then what's driving the sequential assuming there is a 13 week in Q2, what's driving the growth there?

Sure.

Continued expansion in Fab 36 and continued investment in strategic programs both in R&D and in sales and marketing to continue to capture customers, expand the market, et cetera.

Then within your spending profile, you've talked a little bit about how charter has flexible capacity and you have some leeway there in terms of your capacity growth depending on the demand and order environment, can you talk about what you are spending for charter to enable process reporting and the design work that's happening for charter processes, et cetera?

No, we don't disclose that type of information.

Ok.

Then can you give, on the capacity front, can you talk about -- in the past, you've talked about unit processor capacity for 2007, I think in the range of like 65 to 70 million units.

Are you still in that range today looking forward?

Can you give us an update there?

It is Dirk here.

In the past, I think what we've communicated is unit capability in the 2006 time frame in the 60 millionish area and intention to go more to the 80 millionish area next year and 100 million the year before that.

That guidance stands.

We still have flexibility to push and pull on that.

Of course, [it's die sized dependent] but that's still a good guidepost at this moment in time.

Ok.

Is charter still representing the five and 15 million units within those two years as far as the flexibility?

Yes.

It is going to be 10%-ish of that kind of number.

Upwards of 20.

It kind of depends on a given quarter.

Ok.

Thanks a lot.

Last question is coming from Mr. John Lau with Jefferies & Co. Please go ahead.

Great.

Thanks for taking my question.

I've wanted to ask you again, Henri, Dirk, circling back to servers, there has been a lot of focus there and with the competition on the [woodset] launch, have you noticed any recent pricing action that has been a little more aggressive, can you give us some color as to how that is entering the market especially in the area that you're very strong in?

Thank you.

Yes, John, I mean it is a little bit early to see what [woodcrest] is going to be.

Because frankly, there's not that many out there.

It has always been a very competitive environment in the two piece phase.

We're seeing a lot of growth in the four piece phase.

Which is very good for us given the scalability of our platform.

And I expect that--- that of course the competition will want to be very aggressive in that market.

But, as you know, those are bids driven type of pricing environment and so it is very difficult to derive from a few days of availability any sort of trend.

Ok.

And as a follow-up, Henri.

Can you characterize what your ratio split is between 2P and 4P right now?

I really don't want to give you that level of granularity.

But I'm happy to tell you that the 4 piece base is growing nicely and we'll make sure it continues that way.

Thank you.

Ok that, will conclude the call.

Thank you very much for your participation.

Thank you.

That does conclude today's AMD second quarter earnings conference call.

You may disconnect your lines at this time, and have a wonderful evening.