超微半導體 (AMD) 2004 Q1 法說會逐字稿

Ladies and gentlemen, thank you for standing by.

Welcome to AMD's First Quarter 2004 Earnings Conference Call.

At this time, all participants are in listen-only mode.

Later, we will conduct a question and answer session.

Instructions will be given at that time. [Operator Instructions]

I would now like to turn the conference over to our host, AMD's Director of Investor Relations, Mr. Michael Haase.

Please go ahead, sir.

Thank you very much.

Welcome to AMD's First Quarter Earnings Conference Call.

The format of the call today will include prepared comments, followed by Q&A.

The participants are Hector Ruiz, our President and CEO, Bob Rivet, our CFO, and Henri Richard, our SVP of Worldwide Sales and Marketing.

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

The replay number is 800-475-6701, and outside the United States, the number is 320-365-3844.

The access code is 727052 for both of those numbers.

The telephone replay will be available for the next 10 days, starting at approximately 8:30 p.m.

Pacific Time tonight.

Before we begin the call, I would like to caution everyone that we will be forward-looking statements about management's goals, plans, and expectations.

As you know, the semiconductor industry is generally volatile.

Our product and process technology development projects and our manufacturing processes are complex.

Current worldwide economic and industry conditions make it especially difficult to forecast product demand at this time.

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 risk factors and our business.

You will find detailed discussions in our most recent SEC filings, including AMD's annual report on form 10K.

With that, I'll turn it over to Hector Ruiz.

Thank you, Mike, and thank you, everyone, for joining us today.

The first quarter of 2004 was a high water mark for AMD.

It was a quarter that produced a record top line for the quarter.

Impressive any quarter, but remarkable in the first quarter of the calendar year.

As important, it was a quarter on which we delivered on a very important promise to ourselves -- to make money.

This was our second consecutive quarter of operating profits, and a highly important milestone in our march towards delivering consistent operating profits each and every quarter, year in and year out.

It was, as I have mentioned before [in this call], an example of what we call ``the new AMD,'' and we're very proud of the progress that we're making with our strategy of customer-centric innovation, our continued world class design and manufacturing performance, and our strong discipline of operational flexibility.

Our Spansion flash memory operation is an excellent example of this winning combination.

In the first quarter, we enjoyed record revenues in flash, as we continued to expand our market leadership position in the [newer] flash memory segment.

The products based on our revolutionary MirrorBit technology continue to earn strong customer acceptance.

We continue to strengthen our solid relationships with our top customers around the world, resulting in our especially strong growth in Asia Pacific and the Americas.

In response to this strong customer demand, we continue our aggressive ramp to 110 nanometer technology in each of our leading-edge flash fabs.

We commenced production of 110 nanometer Floating Gate flash product this quarter, the quarter that just ended, and we expect to begin production of our first 110 nanometer MirrorBit product this quarter, the second quarter.

I would like to reinforce that AMD is the first to have commenced [newer flash] production below 130 nanometers.

In the mean time, it is important to remember that while in only the third quarter of Spansion's formal business integration efforts, we again delivered strong quarter to quarter growth, and the first operating profit in our memory group since we combined the operations only nine months ago.

The story in our Spansion operation is very straightforward -- we are expanding both the breadth and depth of our customer base.

We are expanding our capacity with new technology to meet the growing demand, and we continue to work overtime to capture the efficiencies of our new global operations.

In our microprocessor business, we again generated strong profits on solid execution and across our complete portfolio of brands.

Our customers continue to show a strong interest in the unique value proposition of our AMD64 technology.

The value proposition delivers only through the combination of a 64-bit instruction set, integrated memory controllers, and HyperTransport technologies.

AMD64 is also the only way to get the benefit of enhanced virus protection with the X86 software ecosystem, which will be enabled by the upcoming Microsoft Windows XP Service Pack II, and it's an important additional benefit, as customers consider replacing systems in the coming months.

We now count over 1,000 partners in our growing ecosystem.

AMD64 unit shipments more than doubled quarter to quarter, powered by growing acceptance in the enterprise segment.

And our commitment to deliver a pervasive 64-bit computing continues, with the early success of our new Athlon 64 mobile processor, which we launched in this past quarter.

Systems based on our AMD Opteron processor family are seeing a strong customer adoption in financial services, web infrastructure, pharmaceuticals, petrochemicals, and digital content creation segments.

And our new low and mid-power AMD Opteron EE and HE processors will continue to drive the Opteron footprint into high-performance for web server and workstation solutions.

With this product, we're introducing for the first time 64-bit performance in this power [envelope] and it is becoming increasingly clear that the 32-bit only server segment is rapidly approaching obsolescence.

The Opteron 64-FX franchise continues to enjoy a strong acceptance among gamers and PC enthusiasts, and with the introduction of the Athlon 64 FX-53, we have once again set the standard for the definitive cinematic computing experience.

We're very pleased to report that in the first quarter, the AMD64 processor family received awards from Gartner, at their Systems Builder Summit, from ``PC World,'' ``PC'' magazine, and ``PC Professional,'' which recognized AMD for our revolutionary cool and quiet technology.

AMD's 64 technology has earned over 45 awards worldwide during its first year of existence, and we're very proud to be recognized as the true technology leader.

In fact, across our brand portfolio, we saw healthy sales growth worldwide, including record sales in Latin America and continued strength in Asia.

We're very pleased that Founder Group, one of China's premiere IT firms, has joined the list of over 80 customers building Athlon 64 processor-based desktop systems.

Our ASP [projectory] continued upward for the fourth consecutive quarter, a reflection of a strong brand discipline, a differentiated value proposition, and solid execution.

Our manufacturing on AMD64 is on time; yields are solid.

We continue our deep ramp to 90 nanometer technology, remaining on track for volume production in the second quarter of 2004 and delivering products for revenue in the third quarter of this year.

Our customer-centric innovation approach is working, as we present a superior alternative to the competition in the markets we choose to serve.

In manufacturing, AMD continues to demonstrate world-class operating performance in each of our core businesses.

And we remain committed to the discipline of operational flexibility across and throughout the organization.

To summarize, we're seeing increased confidence among our growing portfolio of customers about AMD and our growing leadership profile, and we're looking forward to demonstrating that leadership throughout the course of the coming year and for many years to come.

At this point, I would like to ask Bob to review in more financial detail the results of our recent quarter and provide some comments regarding the current outlook.

Thanks, Hector.

As detailed in our press release earlier this afternoon, our first quarter sales were $1.236b, up over $500m or 73% compared to the first quarter of last year, and up 3% compared to the fourth quarter of 2003.

This is an all-time quarterly sales record for AMD.

Top line highlights include, one, solid sales in a seasonally down quarter for both our microprocessor-- computational product groups, excuse me, or CPG, and our memory group.

Two, especially strong sales in Asia Pacific and Latin America in both businesses.

Three, increased ASPs in CPG and our memory group.

And four, increased unit shipments in our key new product families, more than doubling our AMD64 unit volumes and continued penetration of products based upon our MirrorBit flash technology.

The combination of top line growth and our operational flexibility measures resulted in a profitable quarter, with net income at $45m, or 12 cents per diluted share for the first quarter.

Operating income was $61m in the quarter, an improvement of $15m from the fourth quarter, representing a 50% sell-through on incremental sales.

Compared to the first quarter of last year, operating income has improved by $187m, on an increase of $522m of sales, a nice turnaround.

Gross margin was 38% for the quarter, up two percentage points compared to the fourth quarter of 2003 and a seven percentage point improvement from the first quarter of last year.

First quarter gross margin dollars increased by almost $41m, a 9% increase from the fourth quarter of 2003.

The gross margin improvement was largely driven by a richer product mix, and higher ASPs in both of our major business lines.

Marketing, general, and administrative spending was $180m in the quarter, up 11% compared to the fourth quarter, due mainly to the following three items.

One, increased marketing expenses associated with our ongoing brand efforts and incremental investments in sales and marketing in our emerging markets.

Two, Fab 36 start-up expenses.

And three, other strategic administrative expenses.

Roughly a third of the increase quarter to quarter were one-time events.

Cash flow from operations was a positive $206m for the quarter, which represents our third consecutive quarter of delivering positive cash flow from operations.

EBITDA was $373m, a 4% improvement over our fourth quarter performance.

Now I'll switch to a business overview for the quarter.

First I'll start with our flash memory group.

In the first quarter, sales grew to $628m, up 188% from the first quarter of 2003 and up 11% over the fourth quarter.

The sales growth was due to strong performance in the wireless and embedded markets and increased shipments of products based on MirrorBit technology.

I'm pleased to report that the flash memory group was profitable in the first quarter, with an operating income of $14m.

This is a strong reflection of a very solid progress in rapidly integrating AMD's and Fujitsu's flash memory businesses.

Gross margins improved in the quarter due to increased sales, higher ASPs, and the initial realization of cost synergies.

The memory group reported a $17m bottom line improvement over the fourth quarter on $61m of incremental sales, or a 27% [fall through].

First quarter sales in CPG were $571m, a 22% increase over the same period a year ago, and slight decrease from the fourth quarter.

In line with normal seasonality, unit shipments were down.

ASPs improved for the fourth quarter, though, and were enhanced by continued penetration in the enterprise market.

With a richer product mix, the CPG business unit increased its operating income in the quarter to $67m, a 7% increase from the fourth quarter.

Turning to the balance sheet, a few highlights.

Cash balances ended of the first quarter at $1.3b, about flat from the fourth quarter.

We ended the first quarter with inventory levels at $693m, roughly flat from the fourth quarter level.

Capital expenditures were $202m in the first quarter, compared to $164m in the fourth.

Now let's talk about the outlook.

AMD's current outlook for the second quarter of 2004 is based on the following projections -- in our memory group, we expect sales to increase modestly.

For our computation products group, we expect sales to decline modestly, in line with industry seasonal patterns.

In aggregate, AMD expects seasonal patterns to prevail, and sales will be approximately flat.

For your modeling purposes, please consider the following.

We still expect fiscal 2004 capital expenditures to be around $1.5b, and we're confident they will be largely funded by cash flow from operations.

We recorded a first quarter tax rate of 5%, and we continue to anticipate that our quarterly tax rate for fiscal 2004 to be within the zero to 10% range.

Given our memory group sales expectations, to be modestly up, we expect the minority interest line to be approaching zero in the second quarter.

Once again this quarter, approximately 55 million shares were included in our diluted share calculations to reflect the shares from our $402m convertible.

Likewise, the diluted share count calculation does not include the interest expense associated with this convertible.

Quarterly earnings of nine cents per share will trigger the inclusion of these shares.

Similarly, the 21 million shares from our $500 million convertible will be included when our quarterly earnings reach or exceed or 27 cents per share.

In summary, we are very pleased with the progress in the first quarter, as we increased sales and profits in a seasonally down quarter and saw our memory group reach profitability.

Looking ahead, we will continue to tightly manage our cost structure by completing rapid transition to our new technology nodes in both our major business units.

In short, we will continue to do what we said we're going to do, an important operating principle in the new AMD.

And with that, I'll turn it back to Hector.

Thank you, Bob.

We have begun this new year with substantial momentum in each of our major businesses.

In our AMD64 and MirrorBit technologies, we have designed truly customer-centric innovations that have begun to reset the competitive balance of their respective markets.

And each has a lot of market headroom in which to expand.

With continued market help, we can be quite bullish about prospects for increased demand over the coming quarters and years, which is why I would like to take the last few minutes of my prepared comments to discuss what I consider an under-recognized capability of AMD, our design and manufacturing leadership, because in addition to a clearly superior portfolio of products and healthy customer demand, leadership in this industry is characterized by sustained mastery of highly complex and evolving manufacturing.

And AMD has built some important, arguably distinctive, design and manufacturing capabilities, which will allow us to continue to thrive as our new technologies come into full maturity.

You don't start with revolutionary designs.

AMD takes great pride in creating elegant customer-centric designs that are simpler, more in line with the specific needs of our customers and their customers.

But beyond this, we design for manufacturability.

For instance, because it is largely derived from prior generations of the technology, our AMD64 technology takes advantage of the great legacy of manufacturing knowledge, developed over years of advanced logic manufacturing techniques.

Our MirrorBit technology is revolutionary in another way.

It is actually a very elegant design in its simplicity, requiring few process steps and resulting in a product that is both superior in its performance and easier and therefore more cost-effective to manufacture.

Our product management discipline has been well-regarded in the industry, both in the context of our performance against leading industry benchmarks and as important, in the feedback from our world-class customers and partners.

But what only a few in the industry understand is the leadership position AMD has quietly established in the process automation arena.

AMD's automated precision manufacturing, or APM, is our patented suite of more than 250 leading-edge fab automation and optimization technologies used to reduce time to yield on new technologies, improve wafer yields, and decrease manufacturing costs.

With APM, each generation of flash memory has ramped to mature yields better and faster than prior generations.

Our best time to yield in our flash business has been in converting to 130 nanometers, which is now fully accomplished in both of our leading-edge fabs.

And it is important to note that we're already ahead of the ramp in our conversion to 110 nanometers in Fab 25.

With AMD Opteron, we experienced a 66% improvement in our time to mature yield ramp over the prior generation of technology.

And all indications are that our performance on 90 nanometers will be on the same or better ramp.

APM is one of our sources of competitive advantage, and we continue to evolve it as an important part of our growing leadership in manufacturing.

AMD is gaining widespread recognition as the clear industry leader.

In leading the computer industry to pervasive 64-bit computing, in establishing MirrorBit as the new industry standard in flash memory, and in empowering a new generation of tech-capable consumers around the world.

We have the design and manufacturing expertise to deliver on these and many other promises in the years ahead.

As I said before, we are the new AMD and we're very proud of the progress that we have made, and of the leadership opportunities that we have created with and for our customers.

I would like to again thank the thousands of AMD employees around the world for all the hard work to get us back to profitability in these past two quarters.

I thank you for your attention and I'd like now to turn it back to Mike Haase for the Q&A.

Operator, we're ready to start the Q&A session, please.

[Operator Instructions] One moment, please, for the first question, which is from the line of Tim Luke with Lehman Brothers.

Please go ahead.

Hi to you, and congratulations on the quarter.

I was wondering, with respect to the gross margins in the quarter improving, whether you could give some color on how you see the margin outlook developing and how you see it, longer term, as you transition to 90 nanometer?

Thank you.

Only upward from here.

In both businesses we see actually improvement, as we transition-- as we continue to complete the transition to 110 nanometer in flash and as MirrorBit continues to gain more traction as a bigger part of the portfolio, we'll see significant gains in gross margin in our memory business and in microprocessor, really, the story is the same.

As we transition to 90 nanometer, we'll see improvement from that, and also as we continue to ramp AMD64.

Can you give us any sort of sense of the range to which that might apply?

Could we, on flattish sales, anticipate a similar jump to the one we've just seen in terms of the gross margin, in terms of the order of magnitude of the improvement?

Well, it really depends on how you model the mix.

Obviously the flash business is quite a bit lower than the microprocessor business, so-- and kind of think of it as almost 50% to one, so it kind of depends on that and really, what mix you assume for our business model.

That's about all the color I'll give you.

Thank you very much.

And our next question is from the line of [Mark Edelstrong] with Morgan Stanley.

Please go ahead.

Nice job on the quarter, guys.

I had two questions.

First one really just related to the incremental profitability you can drive from the operations in flash -- maybe just give us a sense as to how far along you think you are in terms of wringing out some of the efficiencies in the merged business and where that could go from here, all other things considered equal, in the flash business for you?

Well Mark, there are a number of things that are going to help us improve, as Bob pointed out.

In no particular order, frankly, it's just the fact that, you know, the synergies of the two institutions combined are just now beginning to be realized.

As Bob mentioned in his comments, we just saw the beginning of the impact in this past quarter.

We expect that to continue between now and the end of the year, to continue to improve.

The second part is we're aggressively moving to 110 nanometer technology in our flash product line.

That would also have an impact, and late in the year, as we begin 90 nanometers, that will also continue that trend to improve.

However, probably one of the things that's very critical here to get a sense for is that the conversion to MirrorBit is going to be a very significant in our ability to improve profitability in that business.

As I mentioned in my prepared remarks, MirrorBit is an incredibly powerful technology, it's simple, elegant, easy to manufacture, in relative terms, because nothing is easy in this business.

But-- and so we're really looking forward to that.

The combination of all these things give us confidence that quarter after quarter, we'll see an improvement in our margins in flash.

I guess, Hector, is there a way just to try to calibrate how much you can get from each of those variables?

That probably is-- any way we calculate it today will probably change tomorrow, so I can tell you they are good, they're all roughly, you know, equal in contributions, but they're all positive, and that's about all I can tell you right now.

OK, thanks for that, and just one last follow-up on the processor business, can you just tell us what happened to ASPs in the 32-bit processor family?

Were those flat, or did those actually increase again?

Mark, this Henri.

The ASP increased slightly in the 32-bit family.

Great, thank you, guys.

Thank you.

And our next question is from the line of Andrew Root with Goldman Sachs.

Please go ahead.

Hi, thanks a lot.

Just to continue on the topic of mix, your 64-bit unit shipments doubled in the quarter.

Curious if you're expecting similar momentum, as those product families ramp, in Q2, as part of your guidance?

We have gained a lot of momentum, and we're happy with it.

We expect that momentum to continue.

In terms of dollar terms, as I mentioned at the last earnings call, and I'll repeat it on this one, is towards the end of the year, we expect the dollar revenue on AMD64 will switch over, over our K7 architecture.

OK.

And then related to that, as the 64-bit architecture ramps, the- when you launch 90 nanometer Opteron for commercial volume in Q3, do you expect at that time that any other key food chain partners, chipset partners, will come alongside that would be interesting to the industry?

I'm not really sure I understood the question.

Well, do you think that, for example, you will see additional producers of chipsets for Opteron product when 90 nanometer is launched in volume in Q3?

Yeah, we're not going to announce or pre-announce product from our partners, but it's clear that there is a growing momentum in the chipset arena for server solutions based on the AMD64 Opteron processor.

OK, I'll take that as a ``yes'' for the moment.

Thanks.

Our next question is from the line of Michael Masdea with CSFB.

Please go ahead.

Yeah, thanks a lot and congratulations.

If I understood you correctly, you guys are talking about fairly seasonal computing unit demand, but you're continuing to see the ASP work in your favor.

If you can talk about-- if that's right, if you can talk about-- is this just a continued kind of customer perception change, and how much more room do you think you can close the gap with your competitor there, is the first question.

Well, there has been a concerted effort, as we introduce the AMD64 family of products, to continue to penetrate segments where we did not serve before, and most obvious one that stands out is the server space and the workstation space, high-performance computing.

We went from pretty much a year ago being a zero in that space and we're gaining traction, and as you know, with more customers signing up with now, with, you know, Sun, HP, and IBM continuing, and others, we have-- we will continue to move into that space where prices are higher and therefore the mix change will continue to put some upward potential in our ASPs.

Great.

Just a quick follow-up, maybe, on the flash side -- can we have some milestones in terms of where you are in MirrorBit now and where you expect to be, kind of by the end of the year, if you can give those out?

What I can tell you is the growth in MirrorBit is robust, aggressive, and we're moving it as fast as, you know, customers are able to move to that.

And we believe that, you know, roughly, in round numbers, give or take a little bit, towards the end of the year half of our products being manufactured will be MirrorBit, or more.

Thanks a lot.

Our next question is from the line of [Joseph Osha] with Merrill Lynch.

Please go ahead.

Hi, guys.

Yeah, two ones.

First, on the flash side of the business, how-- how close are you to capacity or being capacity-limited, and then I have a follow-up.

In the first quarter, the one that we just ended, our flash capacity was very tight.

We-- you know, we live hand-to-mouth in many, many-- many of our products, but particularly the heaviest challenge was definitely on the low-density part of the equation.

As we go into the second quarter, we're going to improve our ability to make product, but we still expect to be very, very tight.

We believe that the conversion to 110 nanometers will begin to improve that in the third quarter.

In the third quarter?

Yes.

OK.

And then the second one, I know you can't-- not going to give me a number, but as I look at my model, it looks to me like Opteron plus Athlon 64 in the first quarter was probably in the neighborhood, in unit terms, of about 10% of your volume.

Am I in the right-- excuse me, 5% of your volume.

Am I in the right ZIP code, perhaps?

You're too low, and I'm not going to give you the number, but you're too low.

Oh, OK.

OK.

So 5% is too low?

All right, thanks.

Our next question is from the line of Hans Mosesmann with Schwab Soundview.

Please go ahead.

Yes, thank you.

A couple of questions on-- can you give us a sense of the Athlon 64/Opteron ramp -- is it coming along as expected, a little better than expected, and the follow on is on flash -- if you can comment on the level of allocation, if you are on allocation, of this product for the next couple quarters?

Thanks.

Yeah, we've modeled our-- the balance between our AMD64 product, which is our K8 architecture, and our Athlon XP product, which right now is mostly known as our K7 architecture, based on what customers wanted-- want from us.

For example, if a customer wants a- the best value you can get, at a given performance, for a 32-bit system only, it is hard to beat Athlon XP today.

And so we've dialed that, as we-- the quarter evolved, and we have managed to stay on plan, relative to the ramp that we can-- we're happy to have on AMD64.

We expect that to accelerate, as in the second half of the year, the software ecosystem begins to get stronger and Microsoft will begin to release the client OS that will address the needs of the 64 system, so we expect that to accelerate.

But we're very fortunate to have now the broadest portfolio in X86 architecture we've ever had in the history of the company, when you go from the value space, mainstream, 32/64 client, server, workstation, and we are dialing in the proper means, based on our customer needs and demands.

And on the flash side, as I said before, we're tight in manufacturing.

We're having a tremendous challenge on the very low density part of the spectrum on flash.

We expect that very tight situation to continue in the second quarter and begin to be alleviated in the third quarter, as 110 nanometer takes over.

OK, thanks a lot.

Our next question is from the line of Krishna Shankar with JMP Securities.

Please go ahead.

Can you give us an indication of what channels and markets seem to be most receptive to adopting the Athlon 64 architecture and can you also give us an update on your efforts in the notebook PC market?

Sure.

In terms of channels, we had, again, a record quarter in the distribution channel, which indicates continued strong adoption by enthusiasts, PC enthusiasts, gamers, and the small and medium businesses.

And that's a statement for the client side.

As we indicated in our remarks, on the server side, the quarter was characterized by increasing adoption of the Opteron processor in the enterprise, outside of its traditional high performance computing segment, and into true computational needs that are more-- considered mainstream.

So I'd say across the board.

As far as the notebook, as you may know, we introduced in the first quarter our Athlon 64 mobile processor and I'm very pleased with the initial response that we've got in the marketplace.

As you know, the most dynamic portion of the notebook market segment is the desktop replacement, and the AMD Athlon 64 mobile processor brings an incredible level of performance for that particular type of usage, so we're seeing strong demand in that segment.

OK.

And on the flash market, given that you're running almost at full capacity, what steps can you take to improve your mix of units so that you get to significantly higher levels of profitability before the 110 nanometer technology gives you effectively more capacity?

Well, the steps-- it's not a question of dialing a different product mix.

It's a question of what our customer demand is, and we're working with our customers to look at the-- you know, their needs and particularly in the wireless space, where the need for higher density continues to be strong and moving up, and we have a very strong partnership with many of the-- all of the key suppliers in the mobile space.

And working with them, the density of our products continues to increase in that particular space.

So we're spending a lot of time and energy working with them to be able to- to make sure that we meet their needs.

And in addition to that, on the embedded side of the house, we're working to try to move people also to MirrorBit, because it's a much more effective technology, better for the customer, better reliability, better cost, et cetera.

That will also lead to improved margins.

And we're putting in place expansions in our flash factory in Japan, which is called JB 3, which is one of the factories that came from Fujitsu, in order to add more capability at these leading-edge type of technologies, both for Floating Gate and MirrorBit.

We expect all those things will continue to modestly increase pricing.

One thing I just want to re-emphasize is that despite the capacity constraints being tight, there's still a pretty heavy duty competition out there in the marketplace, and although prices are firming, they're definitely not a rocket going up.

Henri?

Just want to underline that in the first quarter, we had the improvement in both average megabits shipped and also ASP, so there is already a trend towards an enrichment of the product mix.

Thank you.

Our next question is from the line of Ben Lynch with Deutsche Bank.

Please go ahead.

First question, the CPU revenues declined, EBIT went up, you know, there's been some concern that you guys were maybe having some problems at 130 nanometer yields, which of course-- well, which you have denied.

How should I interpret this revenue-EBIT divergence?

Is it, you know, AMD64-bit products have, you know, higher margins that 32-bit, or is there something else behind it, and I have a follow-up question, please.

Well, I guess I'll repeat a portion of what I said before.

The introduction of AMD64 has given us the opportunity to penetrate segments that we did not serve before, and those segments have allowed us then to improve our ASP and product mix.

That is a very positive statement and has helped in our ASPs.

The second part, and not-- you know, I'm looking for ways in which I can dispel this belief in the marketplace, is-- we-- our manufacturing is at an all-time high in terms of yields, performance, cycle time.

Benchmarks done by third parties, which are made up of practically every semiconductor manufacturer in the world, indicate that we are the premiere manufacturing site, premiere manufacturing company.

Our customers, as present data to them, they're pleasantly receiving the data that we present to them, because we are very strong.

So 130 nanometer is mature, the yields are outstanding, and as we begin to make the transition to 90 nanometers, which I said in my prepared remarks, we're going to begin within a matter of a few weeks, we already have data that indicates that we will begin 90 nanometers at a higher level of productivity that 130 nanometer, so we're very pleased with manufacturing and that's why I took some additional steps in m prepared remarks to try to convey to you how strong we are in that particular area.

Yeah, thanks.

OK, the second question I had was maybe for Henri -- CPU revenues outgrew Intel a lot, or let's say they ``under-grew less,'' probably about eight percentage points or so.

For the second quarter, you're guiding maybe only 1% to 2% better, it seems.

You know, are you losing sort of momentum in terms of the rate of share that you're gaining, or do you think you're more so just being conservative?

Well Ben, as you know, the second quarter is always a tricky quarter for this industry, in particular in Europe, where we have a very strong presence.

So I would say we're conservative.

Great.

Thank you very much.

Our next question is from the line of John Lau with the Banc of America.

Please go ahead.

Thank you.

Henri, given the rapid adoption of the AMD64 and looking at the forecast for the second half of 2004, are you going to be constrained by capacity at the 130 nanometer node, and what is the percentage of the-- the best guess that you have in terms of what the AMD64 percentage would be at 90 nanometers by the end of the year?

And I have a follow-up, thank you.

Do you want me to answer the demand question?

Yes.

Well, John, from a-- as we can see today, based on our plans and forecasts, I don't foresee any constraints of supply of AMD64-based processors for the second half of the year.

That's the response, from a demand perspective.

As far as the 90 nanometer question?

Yes?

Can you clarify exactly what you want to know there?

Yeah, what percentage of the AMD64s will be at 90 nanometer by the end of the year?

You know, I won't give an exact amount.

I think it'll be significant, because we're ramping the technology fairly aggressively, beginning this quarter.

OK.

OK, and then in terms of fall, looking into the second half of 2004 and into 2005, in terms of the relative growth based on a unit basis, do you believe the notebook market units will be growing faster, or is the server units still the catalyst for 2004?

Well clearly, from a unit perspective, the mobile market will be a catalyst for growth for us.

As I mentioned, we introduced the AMD64 mobile processor in Q1.

That was the first order, and we expect to grow that business very quickly.

From a revenue perspective, we're going to see both the mobile segment and the server segment generate, you know, good revenue growth, obviously, because of the ASP driven in the server segment.

The only thing that I would add to that is that because we're going from an incredibly low level of participation in the server space, our opportunity for growth for us is much higher than what it is for the market.

Great.

Thank you.

Our next question is from the line of Quinn Bolton with Needham & Company.

Please go ahead.

Hi, first question for Bob.

I think in some of your earlier comments at investor conferences this quarter, you talked about a double-digit operating margin target for the flash operation by the end of '04.

Still think that that's a good target?

Yes.

By the fourth quarter, still think that's within reach, and again, that's based on demand is very strong, MirrorBit technology continues to ramp very aggressively in the marketplace and accepted by our customers, and the transition to 110 nanometers now and starting 90 nanometers at the end of the year.

OK.

And then the second question -- can you give us an update on the New Castle product and how that's ramping here into the second quarter?

Yes, the product is on plan.

OK.

Is our update on that.

All our product plans are on plan for the second quarter.

And our next question is from the line of Michael Mcconnell with Pacific Crest Securities.

Please go ahead.

Ah, yes, on the high density flash pricing, is it more a function-- the rather benign pricing we're seeing out there, although it is stabilizing your ability to really get better pricing, is that more a function of your competitors' pricing tactics right now, or one of your competitor's pricing tactics, or more a function of the longer term contracts you have with your customers?

It's the latter.

As you know, a vast majority of our flash business today is with strategic OEMs around the world, with whom we've engaged in multi-year relationships.

Therefore we have a stability in the pricing that's essentially driven by the performance of our product portfolio and our roadmaps.

There is pricing pressure out there, but you know, those type of customers don't switch over simply because there's a lower price suddenly in the market.

And then what is your target by the end of the year for, on the flash side, for 110 nanometer, percentage of total flash by the end of the year, your capacity?

More than 50%.

And Bob, what was the add back to the converts in Q1?

55 million shares, the first convertible, the $402.5m convertible is in the fully diluted share computation.

OK, great.

Thanks.

Our next question is from the line of Chris Stanley with JP Morgan.

Please go ahead.

Hey thanks guys, nice quarter.

Just a couple of clarifications on flash.

You said ASPs go up -- did actual pricing go up in Q1 for flash, i.e., pricing for individual chips?

No, as you know, again, because we're engaged in many long-term relationships, typically Q1 is a quarter where you negotiate pricing for the year, for the fiscal year.

Yeah.

So there was the natural adjustment in those relationships, of the pricing.

But yet we had an ASP increase, which is a testimonial to the growth of our-- you know, the quality of our mix.

Exactly.

And--

-there was a slight decrease in price and an increase in ASP, due to a, you know, significant increase in product mix.

Got it.

And do you expect a similar situation to occur in Q2 and Q3?

Well as I stated, we negotiated most of the long-term relationships in the first quarter, so as far as the large customers, we now have a stable situation.

Clearly for us, we're driven by our customer demand to move more and more of our solutions to MirrorBit, and that, you know, provides us with opportunity to bring, you know, cost-competitive solutions, but at the same time, improve our own profits.

Sure.

And then can you just comment on what the lead times in are in flash right now?

Our lead times in the 12 to 13 weeks range.

And-- so it sounds like you expect those to remain in the same range for Q2 and then hopefully be brought down in Q3, when the capacity comes online?

That's correct, and you know, again, we are ramping 110 nanometer on Floating Gate.

Next quarter we'll ramp 110 nanometer on MirrorBit and the conjunction of these two transitions will bring us in a very favorable position to respond to customer demand in the second half of the year.

And last question -- can you give us any sense of how much capacity you have coming on in Q2 and Q3?

No, I wouldn't give you specifics, but I can tell you that we're moving rapidly towards the-- if you were to look at fourth quarter to fourth quarter, we're-- it's approximately doubling.

I'm sorry, say that again?

If you're looking at fourth quarter to fourth quarter, we're probably doubling, more or less, give or take a little bit.

Great.

Thanks a lot, guys.

Great quarter.

Our next question is from the line of David Wong with A.G. Edwards.

Please go ahead.

Thank you very much.

For your flash sales, can you give us some idea of the distribution amongst areas, major end markets?

Sure.

We're-- clearly our biggest position in the wireless handset market, where we're engaged with most of the key players in the industry, and that represents a little over 60% of our business, but we also have a very strong position in the embedded segment.

And actually, it's important to note that in the first quarter, the growth in the embedded markets was slightly higher than in the wireless market.

And in embedded, can you give us any idea of what the major users of the embedded are?

Well, I'll give you-- you know, consumer electronics and automotive are traditionally very strong segments for AMD.

OK, thank you.

And our next question is from the line of Adam Parker with Sanford Bernstein.

Please go ahead.

You might have said -- I apologize if I missed it -- but when is your plan for 50% crossover on production on 90 nanometer in the processor space?

I didn't say for 90 nanometer; but--

You said production is starting in Q3, so when do you think you'll cross over?

90-- you know, it'll between fourth and first quarter.

OK.

Then just following up on something Bob said, just trying to dive a little bit more into your comment that margins are only sort of upward from here.

You know, it seems like at a high level, the biggest factors are basically depreciation, technology improvements, and then a combination of pricing and mix.

And I just want to talk about the kind of depreciation part a little bit.

Can you help with-- do you think you've really structurally reduced your capital intensity over time, and you know, what's the depreciation impact in Q2 versus Q1, and then sort of longer term, sort of 2005 versus 2004, if you can add any color?

Well, we're-- if you haven't picked up on earlier answers to questions, we are adding capacity.

We're [inaudible] machinery capacity in the memory business in JB 3 in particular.

Right.

That building was only actually half equipped, so we will have more depreciation come on line throughout the year.

Most of that will happen in the second half of the year, from that standpoint.

And you know, we're enjoying leadership position in that marketplace.

We continue to try to distance ourselves from number two.

So we'll continue to add capacity at the appropriate time, which to me will yield a little more depreciation.

So I guess my point is, you know, do you think longer term, you've kind of changed your capital intensity at all, or is that, you know, is that not part of the plan?

We always continue to-- I guess the answer is we always look at it, and figure out where to find the sweet spot, but the memory business is still a very capital-intensive business.

All right, so the second piece of that, then, would be kind of the pricing, or you know, you've talked about the mix being good, but can you maybe talk a little bit about, you know, what you envision for flash pricing over the medium term here, with I guess Intel, Samsung, Renaissance, all wanting to gain more share in [newer] flash in the coming couple of years.

How do you think that's going to plan out, or do you just think your technology is just a cut above and getting better and better?

Well, I think I wouldn't be so bold as giving you a two-year outlook on price in flash.

If I did, maybe I would have your job, but--

That's the last thing you want!

But clearly again, because of the strategic nature of our business, and particularly in the wireless space, you know, that area is really driven by constant innovation and progression in density and complexity of the package.

So, you know, time to market, a competitive roadmap, and high-level quality technology is really what matters.

At the low end, in the lower density, as you pointed out, there are some people that have lost significant market share over the last year that are going to try to regain some of that share and I expect that a lot of the competition may come in that space initially.

So yeah, I guess the--

--so I would say, you know, competitive, tough business at the high end, but again, it's more than price.

As you go down in density, you know, the effect of being cheap and available is probably more of a factor of decision for the customers.

And I'd like to add to that just one thing -- you alluded to it in your question, is that the MirrorBit technology is really unique.

It's incredibly strong, very powerful in terms of cost performance and quality, and I think that will be one significant competitive differentiator.

OK, thanks for the color.

Our next question is from the line of Clark Westmont with Smith Barney.

Please go ahead.

Hi, most of my questions have been answered.

I'm just wondering on the-- in terms of any roadmap or milestones you've put out there for Athlon 64, can you tell us how you're tracking to those or give us any milestones for the year in terms of unit volumes or anything like that?

Well Mark, as Hector pointed out, we've always said that we were expecting the crossover point to be in the fourth quarter of 2004, and we're maintaining that guidance.

OK.

And any mid-year-- I think you had said at one point, over a million units cumulative, or a million and a half?

I'm trying to nail down which one of those numbers was the bogey, or was there such a bogey?

Well, the market is determining that, but we are-- we are-- our capability to ramp is on plan, we're happy with where we are, and we expect, again, on the longer term view, to cross over by the end of the year.

And I think there was an earlier question, about whether the margins on the 64-bit are more than the 32-bit.

Can you comment on that?

The recent [data] right now is that we are so heavily influenced by the fact that the early introduction of AMD64 was in April of last year, a year ago, and it was not in the server/workstation/high performance computer, where margins are, you know, higher than the 32-bit performance we had, which was mostly in the consumer client space.

For that reason, that continues right now.

However, part of the client that I think is important for you to know that is doing extremely well is in the Athlon 64-FX, which is really the highest-performing client product that you can get, period, in both 32-bit and 64-bit.

And those margins are pretty good, too.

Thanks.

Operator, we're going to take two more questions, please.

Yes, sir.

Our next question is from the line of [Alan Michon] with CIBC.

Please go ahead.

Hi.

With your expectation that revenue will be flat next quarter and the slight shift towards memory, we would tend to expect that gross margin would decline slightly, going forward.

Is there anything going on, on the ASP side, or maybe on the manufacturing side, that would cause that not to happen?

You've decoded it pretty correctly.

We're trying our darnedest, of course, to offset that kind of mix shift issue that takes place, but nothing of significance that can happen in a 90-day window of time to overcome the two-for-one issue between the microprocessor margin and the flash margin.

OK.

And I know we've beaten flash margins to death, but just kind of in simple terms, do we get additional margin boost on the flash side from, you know, just a higher volume, or are we dependent on pricing improving at this point, or you know, a shift towards MirrorBit or higher density, et cetera?

Well, higher volume always helps amortize the cost, because it's a very fixed-cost business, so that's one.

Any increase in MirrorBit is a positive.

So-- and obviously ASP is a positive, so you know, we have a lot of levers that are moving in the right direction, it's just a matter of how quickly they gain traction and move through the system.

Great.

Thank you.

Our next question is from the line of Michael Cohen with Pacific America Securities.

Please go ahead.

Yeah, congratulations on the quarter.

Hector, with the recently announced partnership between Sun Microsystems and Microsoft, I was wondering if you'd be willing to share in your words how you think that this partnership is going to affect AMD?

Well, I will give you a personal view, and it's not a reflection of anything from our key partners and customers.

I think it's great for the industry, that these two companies are working together now, and with a strong partnership with Sun, which we are enjoying and we're very thankful and excited about, it can only be positive for us.

OK.

Another question I had for you is with Intel recently announcing that they're going to also have 64-bit extensions on Xeon, I was wondering what differentiators, like an integrated front side BUS and other things, do you believe is a compelling reason why customers should still be choosing, let's say, Opteron over Xeon?

Well, you know, that's a tough question, with the fact that, you know, I personally and I don't know anybody that has seen any of the product yet that is going to compete with Opteron, so-- but given the fact that Opteron is an eighth generation processor, not a seventh generation processor with an added instruction set, but an eighth generation processor with an integrated memory controller, with HyperTransport Technology, and also a-- the inherent capability, because of that, to be able to move into the multi-core space, I believe we have a very strong position, of which the customers will see the value of pretty quickly.

Yeah, Hector, I'd just like to add that, you know, to date, we are winning in many, many segments on 32-bit performance only, so that's clearly, you know, a testament to the fact that the architecture of the product is superior. 64-bit extension is relevant and becomes more and more important by the day, but architecturally, this product is the superior product today because it's a new generation and there's a lot of catching up that needs to be done before you can match the performance of Opteron.

And congratulations on being the first to introduce it as well.

Thanks.

Thanks, Mike, and thank you everyone for participating.

We're going to conclude the call now.

Ladies and gentlemen, that does conclude our conference call for this evening.

Thank you for your participation for using AT&T Executive Teleconference.

You may now disconnect.