超微半導體 (AMD) 2003 Q4 法說會逐字稿

Ladies and gentlemen, thank you for standing by, and welcome to the AMD fourth quarter earnings announcement.

During the presentation, all participants will be in a listen-only mode.

Afterwards, we will conduct a question-and-answer session.

At that time, if you have a question, please press the 1 followed by the 4 on your telephone.

As a reminder this conference call is being recorded Tuesday, January 20, 2004.

I would now like to turn the conference call over to Mr. Mike Hasay, Director of Investor Relations.

Please proceed sir.

Thank you very much.

The format of the call today will include prepared comments followed by Q and A. The participants are Hector Ruiz, our President and CEO, Bob Rivet, our CFO, and Henri Richard our SVP of world-wide sales and marketing.

This call is a live broadcast and will be replayed at AMD.com and StreetEvents.com the telephone number is 800 633-8284.

Outside of the United States, the number is 402-977-9140.

The access code for both is 21180211.

The telephone replay will be available for the next ten days starting at 6:00 p.m. pacific time tonight.

Before we begin the call I would like to caution everyone that we will be making 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 project 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 the annual report on Form 10-K and our third quarter Form 10-Q.

With that, I would now like to introduce Hector De Ruiz, AMD's president and CEO.

Thank you Mike.

And thank you everyone for joining us today.

One year ago at this time, we were focused on some very fundamental changes in our operating model, changes which adjusted would position us for sustainable prosperity.

I'm very pleased to report that this past quarter it's proved that we can deliver on the promise of prosperity.

And it is my full intent, along with all the members of the AMD management team and everyone of 14,000 plus employees, to deliver on the sustainable part of promise as we go forward from here.

I hope that you're beginning to see a pattern of delivering on our promises to our customers, to ourselves, and to our shareholders.

We take that very seriously here and believe this past quarter is yet another in a series of quarters where we did everything we said we would do.

Last quarter I introduced three characteristics that define what I called then, the new AMD.

The winning strategy of customer-centric innovation, we're becoming a strategic enabler to our customers by focusing on what is most important to their success, continued world-class design and manufacturing performance, and a strong discipline of operational flexibility.

The recent quarter is the continuation of our progress in success in each of these three areas.

Our customer centric innovation strategy continues to bear fruit resulting in ever strengthening customer relationships and as a result solid top line growth in each of our core businesses.

In our microprocessor business, we generated strong profits on solid execution and across our brands and geographies.

We are very pleased with the continued strength of our AMD Athlon XP brand.

It is truly the workhorse brand in our processor business and it continues to enjoy a strong success in each of our customers product segments.

Meanwhile, our AMD Athlon 64 product family continues to gain momentum, especially with our top tier 1 customers.

In the past quarter alone best of systems, based on the AMD Athlon 64 processor, became available from a variety of customers, including H-P, Fujitsu Siemens, Fujitsu Packard-Bell, and E-machines.

Our AMD Athlon 64 FX product continued to gain share with a sophisticated gaming audience for which it was targeted.

We were honored that the Athlon 64 FX processor earned the reader's choice award from Tom's Hardware guide for best innovation in CPUs for 2003.

Finally, our AMD Opteron family continues to make strong inroads with both top tier customers and their most important commercial accounts.

For example, Fujitsu Siemens computers announced its sell to V 810 work station, the first dual-processor work station based on AMD Opteron processor 200 series.

In addition, in November we were thrilled to announce a comprehensive strategic alliance with Sun microsystems.

Sun is already shipping blade products based on our AMD Athlon XP mobile 1800 plus processor, and they are on track to ship two and four-way servers based on AMD Opteron in the first half of this year.

Beyond this, Sun's solaris franchise and portfolio of enterprise software relationships bring additional momentum to the AMD Opteron success story.

Our strong progress in the trade and the commercial sector continued, with important customer wins such as Daimler-Chrysler, Qualcomm, Perelli, and Bristol-Myers, Squibb, and others.

We saw healthy sales growth in every geography, including record sales in Latin America and China.

And in summary, our ASP trajectory continued upward, a reflection of a strong brand discipline and solid execution.

In its second quarter of existence, our expansion flash memory brand has extended its marketshare leadership in NOR Flash memory.

Overall, AMD flash sales continue to grow faster than the NOR market in both the embedded and wireless segments.

And customer demand for more feature-rich product continues to drive average densities up, resulting in a richer product mix and higher ASPs.

We continue to gain share in the consumer electronics sector, especially in emerging markets, a function of the strength of the theme that joined us from Fujitsu.

And we expanded our existing leadership position in the wireless segment.

Once again the fourth quarter produced record demand for our Mirror bit technology which continues to be the fastest growing part of our flash portfolio.

All told, the fourth quarter was a record quarter for us in the flash business.

As a reminder, only two quarters into the integration efforts the progress we have made is rather phenomenal and the fact that we came very close to making money in the quarter is proof of our great progress.

Rounding out our portfolio the personal connectivity and solutions group experienced strong sales growth in the quarter.

In fact, in its first full quarter AMD, our AMD Geode brand achieved near record sales, a testimonial to the strength of the AMD brand.

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

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

In flash, we've successfully qualified 110-nanometer floating gate technology for production as we continue our aggressive transition to 110-nanometer in both our fab 25 in Austin, Texas, and JV 3 in Japan.

This, coupled with the growing adoption of our MirrorBit technology will enable continued capacity growth to meet expanded customer demand.

In processors, our manufacturing ramp on AMD 64 is progressing to plan.

Yields are solid.

And we are capable of meeting customer demand as it continues to scale across our entire AMD 64 family.

We continue our steep ramp to 90-nanometer technology and remain on track for production wafer start in the first half of this year.

Finally, as promised, AMD broke ground on a new 300-millimeter manufacturing facility, AMD fab 36, located in Dresden, Germany, and adjacent to our award winning AMD fab 30.

AMD fab 36 is expected to be in volume production in 2006.

While the fourth quarter was clear evidence of both strong top-line performance and solid manufacturing execution, we continued to drive the principles of operational flexibility across and throughout our organization.

We're now the clear market share leader in NOR Flash worldwide, we're driving the industry to pervasive 64-bit computing and our immediate success with the AMD Geode family is the confirmation of the strength of the AMD brand.

To summarize, we're seeing increased confidence among our growing portfolio customers about AMD and our growing leadership profile.

And as Bob will report, the results today have been very promising.

And at this point I would like to ask Bob to review our fourth quarter and year-end financial results in addition to the outlook.

Thanks, Hector.

As detailed in our press release earlier this afternoon, our fourth quarter sales were $1.206 billion dollars, up 76% compared to last year's fourth quarter, and up 26% compared to third quarter of 2003.

Top line growth was a reflection of increased demand in each of our major businesses across all geographic regions, highlighted by record sales in China and Latin America.

The combination of our top line growth and our operational flexibility measures, resulted in a profitable fourth quarter.

In the fourth quarter our net income was $43 million, or 12 cents per share, included in our fourth quarter results is the positive impact of $14 million or 3 cents per share due to an adjustment to the previously recorded restructuring charges and purchase accounting related to the FASL LLC transaction.

Operating income was $46 million in the quarter, an improvement of $76 million from the third quarter.

Gross margin was 35% for the quarter, up 1 percentage point compared to the third quarter.

Fourth quarter gross margin dollars increased by almost $100 million which represents a 40% fall-through on the gross margin line.

Research and development spending was $227 million for the quarter, up 6% from the prior quarter.

A large portion of the sequential growth was driven by initial fab 36 costs and integrating the new Geode engineering team.

Marketing, general, and administrative spending was $162 million for the quarter, up 8% compared to the third quarter, in line with seasonal merchandising activities.

We generated $389 million of positive cash flow from operations this quarter, the second quarter in a row we generated positive free cash flow.

We grew EBITDA in the quarter to $358 million at 29% improvement over our third quarter performance.

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

First I'll start with computation product groups, or CPG.

Fourth quarter sales on CPG were $581 million a 38% increase over the same period a year ago and a 15% increase over the third quarter.

Increased sales were due to solid penetration with our largest OEM customers in the commercial market, and an improved product mix.

We shipped more units and improved our ASP's significantly compared to the third quarter.

As a result, the CPG business unit improved its profitability in the quarter and made $62 million of operating income.

Profit fall-through on incremental sales was excellent at 56%.

Now moving on to our memory group.

In the fourth quarter, memory sales grew to $566 million, up 161% compared to fourth quarter of 2002, and 34% over the third quarter.

In an overall strong market this performance was primarily driven by denser mix, stronger M C P sales, and accelerating MirrorBit sales.

MirrorBit sales continue to be a larger percentage of our flash memory business.

Total unit shipments increased by approximately 16% and average density increased by approximately 19% compared to the third quarter.

The memory group recorded a small operating loss of $3 million in the quarter, a reflection of solid progress in rapidly integrating Fujitsu and AMD original flash business'.

This represents a $46 million bottom line improvement over the third quarter on $142 million of incremental sales, or a 32% fall-through.

We have seen some progress on cost reductions in the quarter but offsetting these gains are ramping factory costs to increase output, a higher percentage of M C P sales, and end-cost increases due to weakening dollar.

We anticipate improved margin expansion and as we capture the cost synergies we outlined previously, and move to the next technology node in each of our flash factories.

Turning to the balance sheet, a few highlights.

Cash balance has ended the fourth quarter at $1.313 billion, up $237 million from last quarter.

Capital expenditures were $164 million in the fourth quarter compared to $138 million in the third quarter.

Now I'd like to summarize a few comments about the year 2003.

For the fiscal year 2003, AMD sales were $3.5 billion, up 30% from 2002.

Annual sales for CPG were $1.96 billion up 12% from 2002.

Flash memory sales concluded 2003 at $1.42 billion, up 92% from 2002.

AMD's operating loss for the year was narrowed to $233 million, roughly a billion dollar improvement from last year, and we made money in the second half of 2003.

In CPG we narrowed our operating loss to $23 million for the year and in flash memory we lost $190 million for the year.

This was great progress in both businesses, particularly in the second half of 2003.

Cash flow from operation was positive for the year at $296 million.

For the year EBITDA was $814 million positive, a $1.3 billion improvement from the prior year.

Capital expenditures for 2003 were $586 million, under our guidance, and 17% less than 2002 capital expenditures of $705 million.

Now I'd like to discuss the outlook.

AMD's current outlook for the first quarter 2004 is based on the following projections.

In our flash memory business, notwithstanding typical seasonal patterns AMD expects sales to be approximately flat for the quarter due to the company's strong position in the market.

For the processor business, AMD expects to be in the range of industry seasonal patterns.

In aggregate, AMD believes the seasonal patterns are expected to prevail and aggregate sales to be down slightly.

For your modeling, please consider the following: In the fourth quarter approximately 55 million shares were included in our diluted share calculation to reflect the shares from our $403 million convertible debt offering assuming conversion.

At a tax rate of 0, quarterly earnings of 9 cents per share would trigger the inclusion of these shares.

Similarly, at a tax rate of 0, the 21 million shares from our $500 million convertible debt would be included when our quarterly earnings reach or exceed 29 cents per share.

Our quarterly tax rate is anticipated to be within the 0 to 10% range throughout 2004.

In 2004 capital expenditures are expected to be around $1.5 billion, including fab 36 building construction and some equipment purchases, equipment purchases for the completion of our flash JV 3 factory in Japan, and the equipment to complete fab 30's transition to 90-nanometer. 2004 depreciation and amortization is expected to be approximately $1.2 billion for the year.

In 2003 we delivered on our promises.

We continued to tightly manage our cost, launch two major products, fully integrated FASL into our operations, grew revenue, and improved our cash position.

Moving forward, we believe we are well positioned for 2004 across all business lines with strong product offerings and a sound financial platform from which to continue to execute.

And in 2004 we anticipate being profitable for the year.

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

Thank you, Bob.

As I mentioned earlier, the fourth quarter is a solid demonstration of our capability as a new company.

Our processor business generated a strong profit based on solid ASP growth, the result of great products, manufacturing excellence, and an improved brand discipline.

Our AMD Athlon 64 and AMD Opteron products continue to attract an expanding list of impressive OEMs, starting with IBM, H-P, Fujitsu, Fujitsu Siemens, and now Sun Microsystems.

From high performance computers to enterprise to consumer we are leading the industry to a future of pervasive 64-bit computing.

And our path to that 64-bit future remains clear.

In an environment that is only becoming increasingly muddy, our performance with expansion brand is just as impressive.

Achieving the leadership position in less than 90 days, while driving integration efficiencies across our worldwide organization.

We are gaining share in both the wireless sector, the key driver of high-density demand, and the embedded sector where our MirrorBit technology is particularly well positioned for success.

MirrorBit technology continues to earn acceptance as the next-generation industry standard.

Our rapid success with the AMD Geode brand both confirms the strength and stability of AMD as the strategic silicone supplier and validates our strategy to drive the X 86 architecture into new market segments as we move toward empowering a new generation of tech-capable consumers in emerging markets.

Looking forward, we are increasingly confident in our capacity to succeed.

Our customer base continues to expand and improve in all our businesses.

And a growing set of customers feels that the strategic contributor to their success.

We entered this new year with the strongest product portfolio in our history and the world-class manufacturing prowess to produce them in volume.

We have our cost structures moving in the direction we want them, and the capacity to leverage even more synergies from our flash business.

We continue to add very high quality talent to our ranks around the globe.

We know that our vision and leadership will be confirmed and in some cases copied by our competitors, but even if they copy us, our customer centric innovation approach will produce a strong customer bond and the insight to lead new ways of opportunity.

We joint the SIA, the WSPS, and the growing number of industry leaders in expressing our optimism for the coming year, and for us the opportunities are clear.

We must establish MirrorBit as the new technology standard in flash memory.

We must establish the strong footprint in the enterprise with AMD 64, and we must empower new technology users in emerging markets.

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

Before I close, I would like to once again thank the thousands of AMD employees for all the hard work to get us back to profitability in this past quarter.

We entered the new year poised and ready to deliver on the promise of sustainable prosperity for our customers, our shareholders, and ourselves.

Thank you for your attention, and now I would like to turn it back to Mike for the Q and A.

Operator, let's get started with the Q and A.

Thank you.

Ladies and gentlemen, if you'd like to register for a question please press the 1 followed by the 4 on your telephone.

You will hear a three-tone prompt to acknowledge your request.

If your question has been answered and you would like to withdraw your registration, please press the 1 followed by the 3.

If you are using a speakerphone please lift your handset before entering your request.

One moment please, for our first question.

Our first question comes from the line of Michael Matibia from CSFB.

Please proceed.

Thanks a lot.

Great job on the execution again guys.

Congratulation.

The first question really is on the wireless side.

It sounds like if I read your comments correctly, that you're acknowledging a seasonal wireless quarter but your guiding for flat.

If that's a correct interpretation can you give us more color if that's more share gain or ASP's or what?

I'm going to ask Andre to comment on this a little bit from the customer perspective, but the market in general being very consumer heavily oriented and the fourth quarter being a very strong seasonal consumer market and wireless handsets being an incredibly strong consumer product, it is the expectation of the industry as a whole that the first quarter will experience downward movement in that particular activity, nothing out of the ordinary, just the traditional first quarter has changed.

As for any specific market segments or comment, Henri, would you?

Thank you, yes Michael.

I believe that we will continue the trend that we set in Q3 and Q4.

It's not particularly new customers just an increased share in existing customer, increase of the ASP and mix.

Great.

And then Bob, maybe a follow-up.

Certainly a little bit less profitability in that segment.

Should we expect profits to come down a little bit as you grow your wireless next quarter and PC's are down, or are you doing more on the cost structure side there?

No, we should continue to have in flash, again to remind you, it's only been 180 days since we actually integrated these business.

We focused on delighting the customer and sales capture.

We're working on cost reductions as we speak and cost synergies so you will see margin expansion as we move forward, on the same sales level.

Great.

Last question for me really on the capacity side.

Do you guys--do you try to pull into capacity, do you feel like you need to because of some bottlenecks in it, if so, given the lead times seem to be stretching out are you having any problems to grow capacity?

We have -- we've started executing on capacity on leading edge technology for flash back in last summer, and the customers alerted us that this was a need coming up and I think we executed that well.

As a matter of fact, today we're meeting customer demands, on MirrorBit, on high density, high performance, flash product.

However, if I take us back to the third quarter of last year, frankly you know, on the lower density older factories, older technologies, we actually shut down our factories for a couple of weeks in the third quarter last year due to an unanticipated slowdown in those particular segments.

Not long after we came back from the shutdown we started experiencing an upside by customers in their expectations, and since then, being in the approximately the August time frame until now, that demand has increased dramatically.

So the places where we are experiencing very tight demand, it's actually in the lower products, lower density, and we are rapidly trying to move as much of the more newer product to the newer factory so we can make space in the older factory to increase that capacity.

We think that that tightness in the low end of the spectrum will probably last us through the second quarter.

Thanks a lot, guys.

Our next question comes from the line of Andrew Root from Goldman Sachs.

Please proceed with your question.

Yes.

Thanks a lot.

Quick question on the 90-nanometer ramp.

If you could talk a little bit about how that will dictate whether you can stay on the very strong operating leverage that you've shown in microprocessors, and I know that you'r scheduled for Q2 production wafer starts and I would assume early Q3 shipments, but is there a critical time during the quarter, are there signs that give you confidence that this ramps going normally, others, not just the microprocessors but in other logic segments have had significant delays in 90-nanometer.

I'm just curious if you think you're free of most of the issues that they've faced.

This is an incredibly difficult technology and challenging and we're very fortunate to be partnering with a company that I believe is the premier leader in technology, meaning IBM, with working collaboratively with them we increase our confidence in 90-nanometers even though the alliance with IBM really is only effectively starts at the 65-nanometer note, but because of that relationship we both have really been able to talk a lot of tough issues, and I am very happy, very pleased, and thrilled with the progress that our teams have made in our 90-nanometer implementation.

The results in our engineering work is very promising, encouraging, and we're very, very bullish and optimistic about what that will do for our business.

Is there a time you'd have to decide to go to production wafers or to go to another tape-out, another rap of the product?

I know some of your competitors have had to go to three or four tape-outs in the 90-nanometer product.

Well, you know, again, I cannot he future is tough to predict but I can tell you our plan right now is based on always having to do some tweaking on product and taking that into account, and that is also factoring into the expectation the shipments of product will occur in the third quarter.

And then, Bob, just the second part of that, if 90-nanometer ramps on time, 50 to 60% incremental operating leverage is what we should look for in M P U's, or does that change at all?

No, that actually would enhance the situation from today's level but obviously as we ramp AMD 64, that currently is a bigger die than the existing Athlon XP, so we need to do that to maintain that 50 to 70% incremental fall-through.

Thanks a lot.

Our next question comes from the line of Tom Thornhill from UBS.

Please proceed with your question.

On the processor side can you comment at all on unit growth or ASP trends in the quarter?

Tom this is Henri, we experienced both unit growth and ASP growth on the trajectory that we had set in Q3.

Okay.

And were -- would ASPs have been up even without shipments of the Athlon 64?

Absolutely.

On the strength of our Athlon XP brand in the fourth quarter we increased ASP on all product segments.

Excellent.

Another question that's probably more accounting than anything else, if I look at the minority interest number, $19.4 million, that's 40% of the loss, about $48 million or so from the flash side, how do I reconcile that with the $3 million operating loss number?

How do I get from one to the other?

Well, it's a little hard.

The business model is a fairly complex one.

Again, you have to remember the FASL business operates as the manufacture selling to the two parents of the distributor, that equates a margin back to the parents.

Obviously we're one of the parents, so that's the piece of the equation you're missing, plus other technology associated cost transfer between the two organization, so it is a little more complex, I think off line we can kind of help with you your modeling, because you can't just divide and come up with the answer.

Yes, I found that out.

Question on incremental margin.

Incremental margin on the processor side fall-through 56% of the operating line, excellent, a little less than I would have anticipated given the revenue growth on the flash side.

What is the prospect of improving that incremental fall-through and what is the a appropriate range that we should be targeting over the next couple of quarters for incremental margin in that business?

A couple of things, and let me do product technology, then Bob will make a comment on the impact of that, and that is when we went live as a new organization with a combination of Fujitsu and AMD in July 1st of last year our emphasis as Bob very correctly pointed out was to make sure we met customer demand, that no customers were affected in the transition, and all our energy was that the merging of these organizations with the serving customers was uppermost in our minds.

Shortly after that we've been looking at the synergies which we just have begun to identify.

The synergies come in two places.

One is the cost efficiency of two organizations that allows you to eliminate some levels of redundancy.

That's beginning to occur and we think we're going to experience that even more fully in the next quarter.

The other part which I believe is the most significant is that both organizations are now aggressively driving the technology changes, moving to 110-nanometers in our factory and beyond 90-nanometers soon there after.

Those things are, when you project them out, bringing about a tremendous opportunity for leverage in -- as revenue growth.

Bob Rivet, would you like to answer?

Yeah, to your ultimate question on how much should we model for incremental fall-through, Hector's really laid out the groundwork here.

As we move through time it will take us a little bit of time to reach these cost synergies, to get to the next technology node, while we also ramp capacity as you probably -- if you heard, we're actually ramping JV 3, filling out the rest of that factory.

There are some cost increases taking place, but bottom line is there's no reason this business shouldn't be delivering 50 cents plus on the dollar of incremental profits as we work through the synergies and the right technology nodes.

So maybe by the second half?

I'd say that might be a little too far out, but I'd say not first quarter, second quarter you'll see some, third quarter.

I think you'll see continued progress each quarter.

Thank you.

Our next question comes from the line of Michael McConnell from Pacific Crest Securities.

Please proceed with your question.

Thank you.

Regarding the launch or the ramp of Athlon 64 looks like we're coming up on -- you guys reducing the cash size by roughly in half to reduce the die size.

Do you plan on any other die size reductions as we get closer to the 90-nanometer version?

At this point in time we don't have any product plans to announce right now

Okay.

On the flash side, I was wondering just the competitive dynamics right now in terms of pricing, if you could give us an update on the high end as well as the low end regarding flash.

Thank you.

On the high end, both for Floating Gate as well as MirrorBit, we have experienced beginning in the last quarter stabilization of pricing, and when it comes to new opportunities and new customers, we are following the demand supply equation and appropriately getting compensated for that.

On the older technology, it's frankly still very competitive.

There are a lot of players around that scrambling for survival, I would say, and frankly that is -- continues to be a very competitive space.

All in all, though, I expect flash products to continue to experience tremendous competition.

We are planning and preparing to have what you would expect on a memory type of product, degredation of pricing and as a result, accompanying the appropriate cost reductions to ensure that we're more than just competitive.

Henri is there anything you need to add on the pricing side?

Our business is very strategic in nature due to our customer base.

As you pointed out Hector it's a very competitive environment, but there's some stability I think what we're seeing in the channel for the opportunistic buyers is clearly a firming up of the pricing, because of the long lead time.

That's all I would add.

Thank you very much.

Our next question comes from the line of Joseph Osha of Merrill Lynch.

Please proceed with your question.

Hi, guys.

Congratulations on progress, progress to profitability here.

Wondering if I can get some qualitative sense in terms of the, you know, percentage of mix either in unit or revenue terms that was comprised by the Kade architecture.

Are we talking say just a couple hundred thousand units here?

We are.

We gave some indication of our plans before.

We are pleased with where we are in our plans.

We're happy with our progress.

We don't feel that we have varied at all significantly from that and expect that to continue.

As far as the mix, you know, we don't give that out, but let me throw out some ideas and help you how we're thinking about it.

That is on the server side, if you segment the market by that way, we go from 0 to practically 100% because we haven't really been in much of a participant on the server side.

If you are going to the high-performance segment of the client we're definitely very strong there and currently with Athlon 64 FX, being a very, very strong player on that side.

When it comes on to the rest of the client space, you're looking now at a transition time.

You're going from 32-bit to 64-bit, as well as from one architecture to another, so it's a little more complex than just the transition and the product.

We are very happy with the progress and we expect that as software becomes more readily available in the second half of the year on the client side, that that will momentum will dramatically increase.

Right now our plans is to have a very aggressive ramp throughout the year for the K 8 architecture.

Okay.

Thank you.

Second question, Bob, let me just save you some time here, rather than going off line with everyone, we'd all like to know how this line item that amounts to about half of reported income works.

I know it's a little complex, but can you perhaps throw some guidelines out that help us to understand how to forecast that item?

You're talking on the memory?

The $19 million add-back, yes.

Well, obviously that's their share of just the FASL part of the business.

Like I said, the piece of the P & L you don't see that we put into the consolidated AMD results of course is the distributor P & L, I'll call it of AMD.

That P & L would obviously have the full value to whatever we sell to end market.

It would have the margin that we've negotiated with FASL to come to each of the parents, and that's reciprocal between Fujitsu and AMD.

Besides the royalty, et cetera.

I'm not sure I can do justice for this on the phone call, to be honest, because there are shared service agreements between the parent and the subsidiary, so there's a lot of different factors that come into play.

In the non FASL piece of the equation, but you see it in the operating income line.

Okay.

But this is a big line item, Bob.

Can you give us some sense as to what, as analysts, we should do here?

Yeah.

The real simple one,Joe, is we anticipate through cost synergies and aggressive movement in technology nodes that we will be sharing in profits with Fujitsu, not losses, and so that number will be getting less, and turn to be a charge for AMD versus a credit, as time goes on.

Okay.

Very well.

I believe that's it.

Thank you very much.

Thank you.

Our next question comes from the line of Quinn Bolton from Oppenheimer & company.

Please proceed with your question.

Couple questions on flash business.

One, you talked about growing shipments of the M C P, or the multi-chip module.

Can you just try and quantify that in terms of percent of shipments?

I'm assuming you get lower margins on that business than you do on just the actual flash die.

Then if you could also talk a little bit about just the end market split.

You talked about strength in wireless, strength in consumer.

Can you give us a sense as to what percentage of sales each of those end markets contributed to the flash business?

Thanks.

First, let me kind of make sure we put this in brackets.

Obviously the flash business quarter on quarter grew significant.

You know, in the 33%, 33, 34% quarter on quarter.

The fastest growing piece of that business was our MirrorBit.

That grew at a significantly higher than the average of the whole business.

Next in line was the -- is the M C P part of the business.

I don't want to give the details on that piece of the business, of what's the pieces, but it did grow more than the average.

Actually, considerably more than the average.

And, yes, you're right, while it carries high sales value, the gross margin is a little bit less on an MCP than I'll call it a standard floating gate NOR technology or MirrorBit technology.

Are there things you can do, particularly insourcing some of the other types of devices, either the Nan flash or the SRAM that go into the multichip packages to improve the profitable of the MCPs?

We have plans that, you know, the details of which, of course, we wouldn't disclose but we have very, very aggressive and quite exciting road maps to do a, I would say somewhat of a unique job of creating a P&L around the MCP business, which is quite different.

We will be able to benefit more than just passing through the cost of additional components.

Okay.

Then just the end market split in the flash?

The only indication I can give you is that the wireless business was strong really around the world, where obviously a lot of the growth in the embedded business came from Asia Pacific, and particularly on the strength of the Fujitsu sales team.

Okay.

Thank you.

Our next question comes from the line of Chris Donnelly from J.P. Morgan.

Please proceed with your question.

Just a couple of quick clarifications.

You said you'll be able to get some cost savings.

Does that mean you think gross margins can be flat in Q1 even if sales are down a little?

That's always a challenge in this business to have gross margins stay the same on lower sales.

It's a possibility.

It kind of depends on the mix of the businesses, but, you know, you can't lose too much altitude in this business being still fairly fixed cost oriented and maintain that without ASP lift or significant cost reductions.

It will take awhile to get to the technology nodes that Hector talked about, in particular in the flash business, and as we talked about in the microprocessor business also, if 90-nanometer comes on in the second half of the year.

So, I guess long answer to your question, a little bit can hold, if it's a lot, it's tough.

That's fine.

Definitely fair.

And then on the flash business if you take a look at say your prices in aggregate for Q1 do you think they will be flat, up, or down for flash overall?

You know, at this point in time, we expect the mix to continue to improve.

We've had a trend of significant and fairly aggressive density uplift on the product and performance and new technologies.

As a result of that, the unit pricing might go up, but the unit number of units may be less.

Is that a yes or a no?

Well, as we said, we expected in our flash revenue, given all the things we're doing, is that we have an opportunity to have flat sales quarter on quarter.

Okay.

Thanks.

Our next question comes from the line of Eric Rothdeutsch, Friedman, Billings, Ramsey.

Please proceed with your question.

Hi, thank you.

Question for Henri.

Henri, in terms of units for flash memory.

What is the highest volume density that you're currently shipping in terms of units, and what is the smallest density you're currently shipping?

I'm not going to give you the highest density volume, I'm just going to tell you our range goes from 1 megabit to 256 megabits and I'm going to remind you that we've got 512 megabit currently on the qualification.

You mentioned that lead times are also very long.

What are lead times currently?

It really varies by product.

But, you know, lead times can be as much as 13 weeks.

Thank you.

Our next question comes from the line of Dan Scovel from Needham & Company.

Please proceed with your question.

Thank you.

Your microprocessor revenue was up 15% quarter to quarter and you said ASPs were, I guess, up significantly.

We can assume, basically, most of that gain was on the ASP side and the lesser of that was on the unit side?

That's correct.

Also you're speaking a little bit vaguely, but I guess you consider yourself--you did reach your goals of shipping hundreds of thousand on the AMD 64 last quarter?

We're pleased with where we are on AMD 64.

And also, Bob, on the accounting side here, what's the adjustment on the convert to calculate the diluted EPS?

We currently included in the current EPS calculation was 55 million additional shares.

You'll see it on the page.

We had about an 18% of dilution between basic and diluted.

And there's a net income adjustment for the interest income there?

Yeah, the model I'd use is about 5.2 million a quarter.

Add-back.

That's what I was looking for.

Thank you.

Our next question comes from the line of Adam Parker Sanford and Bernstein.

Please proceed with your question.

Same issue with the diluted versus basic EPS.

Is the right number $43 million in net income divided by the $4.16 a share?

Because that seems to give 12 cents, not 10.

I don't know where you got 10.

Well, or are you adding back 5 million to the 43 to do the actual calculation?

Seems like the basic and diluted EPS are the same at 12 cents but the diluted share count is 19.8% bigger so I'm just trying to see what I'm not understanding here.

The math is you take $43 million and change divided by the basic to get the basic EPS.

Right.

You take the $43 million of net income, add back the interest charge associated with the convertible, and then do the math.

So you're adding that 5.2 back.

Right.

Okay.

On the share count, just another issue, what was the impact from options here, you know, kind of sequentially, and, you know, in terms of the 19.8% increase in the share?

I know you said the bulk of it came from the -- one of the converts, but can you help at all with what this share count balance was from any options exercised?

If you have my press release in front of you, you can see the transition in basic, $347 million to $357 million.

That's really option activity.

Okay.

So just the 10 million, and the rest is....

Okay.

And then can you add some color to the deferred income on shipments to distributors?

What does that imply about revenue in the next couple of quarters?

Can you just remind me how that works?

The way it works is obviously what we do is we ship product in the appropriate mix that we agree with, with our distributors, obviously that will range from our server base product, AMD 64 Opterons to our lowest based product, so whatever it is.

Then we count it at the end of -- at the end of each quarter we count what's still left.

Whatever is still left we do not record as sales, obviously it's mix-dependent what's still on the shelf.

We clearly see the transition that's taken place and we have more valuable product on the shelf of the distributors as we continue to produce and make progress on the whole portfolio of products.

Okay.

Sorry, one last question.

Can you talk about your plan for op ex going forward?

I think R&D and SG&A were up, about 6 and 8% sequentially.

How should we think about op ex for 2004?

As Hector said, you know, we're optimistic on 2004, and we will toggle those appropriately.

At this point in time I'd play in they're going to grow a little bit.

As we take advantage of market opportunities to capture more customers to deliver the right technologies, et cetera.

So it will grow.

Not as much obviously in the first quarter, but, you know, we'll have to see how it unfolds as time goes on.

Okay.

Thank you.

Our next question comes from the line of Hans Mosesmann from Schwab and SoundView.

Please proceed with your question.

Most of my questions have been answered.

I do have a question regarding flash bit growth for '03.

Do you guys have an expectation of what that was, and what the forecast is for flash growth for you guys in 2004.

We continue to see obviously exponential growth in bits shipped, there's nothing in the demand of high-end devices that's going to change that for '04.

So do you have a number for '03?

Was it 50%? 100%?

Will that bit growth in '04 match that, that you saw in '03?

I don't have the data for the whole year, but for the last quarter it was 30% plus.

That's sequentially or year-over-year?

Sequentially.

Then one last question.

Do you have market share goals for flash and or processors in 2004?

Thanks.

We have, of course, our internal plans, what we think we can accomplish on both flash and processors.

We don't publish that, as you can imagine.

Okay.

Great.

Thank you.

Our next question comes from the line of Mark Edelstone from Morgan Stanley.

Please proceed with your question.

Good afternoon, guys.

First question is on the Athlon 64 and Opteron product line combined.

Obviously you had a good momentum here so far.

Do you have the ability, do you think, to be able to produce and ship as much as a million units here in the first quarter, and then I have a follow-up question for you as well.

Let me just talk about our capability in general terms, Mark.

We currently are running our factory as a mix of 130-nanometer both Athlon XP and next-generation hammer product.

We -- if I -- at the risk of sounding somewhat simplistic here, it's kind of like a knob that we're dialing, and as to how much of that we want to be Athlon XP and how much we want to be Hammer or Reented.

We're basing the market with our customer very carefully as to what they want and need and planning this result from the beginning because we anticipate the beginning of the 90-nanometer conversion to really give us a big boost in our capability for particularly the K-8 hammer architecture, but even at the 130-nanometer I want to make sure it's clear that today, we have the capability of doing if we wanted to, more than a million a quarter, well more than a million a quarter.

Significantly more than a million a quarter, so we're throttling that based on our plans, which are, you know, how to penetrate particular customer segments and accounts and the expectation in the second half of the year, other things will occur relative to software, where we will then be at 90-nanometers and be able to execute a little more leisurely how many units we can produce.

When you tweak the knob, so to speak, should we expect that the peak in the 32-bit Athlon version would peak once your 90-nanometer technology really ramps?

By peaking, I'm trying to understand what you mean by the question.

You mean if the Athlon XP will peak?

Yeah, if we did the 32-bit Athlons, obviously as the Athlon 64 ramps up, at some point that's going to cannibalize the unit opportunity you have overall.

If we make the assumption that your total units continue to grow with the market, at some point we're going to expect the 32 Athlon units to basically peak, and I'm wondering if that happens once the 90-nanometer capacity ramps up or if you see that happening before that.

I think it would be fair to say that as the 90-nanometer ramps up, there is a significant shift toward the KA architecture more and more.

I do want to take the opportunity though to point out that Athlon XP is not a product in our view, it's a brand, and it's intended to address the best value that you can have in today's applications.

That will continue for some time.

We see that as a very powerful brand, has done well for us.

Athlon 64 will become the next generation, the transition to 64-bit, followed by Athlon 64 FX, which is the best you can get.

There's nothing closer to that in performance, then Opteron, of course, is the server and high performance computer end space.

We see this as a four tiering of our brand.

We'll continue to manage that as customers work with us, therefore we're beginning to talk less about the product transition and more about the segment.

Fair enough.

I appreciate that answer, Hector.

One final question, if I could, for Bob, going back again to the minority interest line, I know there's lots of give and take here as you go forward, but as we go into the second half of 2004 should we see that line toggle to basically profit for the combined flash operations, or would you expect to still be paying out minority interest in the way of losses?

No, to me, if we execute to our plans, which we have every indication we will do, and the market continues to behave the way we think it will, that line will actually become a charge for AMD, because the FASL organization will make money, and we will share that 40% of that profit with Fujitsu.

So it'll, it'll--you know, my forecast, if I was doing it for you guys, would be that that number will become smaller approaching zero and will turn positive as the year unfolds.

Great.

Thank you very much, guys.

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

Please proceed with your question.

Hi guys, Hector, congratulations on delivering your commitment of three months ago of turning into the black this quarter on the bottom line.

Are you going to do the same in the March quarter?

Are you going to be profitable in March?

Oh, you know, we are in the situation where revenue drives everything right now.

We've done everything we can, that I believe is appropriate on cost, we've got our factory sufficient, we've done everything we've put in place.

I believe that if the market is strong we certainly are going to do well, and at the current plan, we do not plan to lose money in the first quarter.

So that's it.

Thanks.

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

Please proceed with your question.

Thank you very much.

With your current 90-nanometer plans of starting production wafers somewhere in the middle of the year when do you expect to be able to announce your first 90-nanometer products?

Our current plans to ship product to customers for revenue in the third quarter of this year.

So the announcements will come in the second quarter, or we'll get details of the product in the third quarter?

Frankly, I don't remember the exact dates.

I'd rather not give you one but can get you our plans when we intend to make announcements.

Great.

Thank you.

Our next question comes from line of Hayden Ugan from Susquehanna financial group.

Please proceed with your question.

Wondering what kind of inventory level do you see channel for both of your flash products as well as your CPU's?

On the CPU side we have a policy to keep our channel below four weeks, we stick to that.

In the flash side I will tell you that the channel inventory is depressed right now.

I don't have a data point to give you.

I would say it's down to probably less than twoe weeks in some of the high run-rate parts.

Then on your PC business, on Q4, do you ship any Duron processors?

We shipped very small quantity of Duron, which was really the end of life shipments of this product.

Okay.

So going in Q1 do you expect to have any more Durons?

Irrelevant.

Irrelevant quantities.

Then what is your -- when do you expect to have your first product coming out on 300-millimeter?

Well, we've we expect to be in production in 5-36 in the year 2006, and that about right now, given the challenge in making and forecasting this sort of technology, that's as close as I tell you a date right now.

Final question is on your Geode.

Right now do you have any plan to expand that product line by integrating AMD processor core into that product?

Geode is going to become part of our road map of X 86 products.

We are extending it rapidly, and we'll communicate that as we get things lined up.

This will actually be a road map of low-cost, low-power and in some cases a synthesizable core.

I don't want to go into what cords are going to go into what product but it's going to be a broad range of applications for the embedded and consumer space.

Great.

Thank you.

Operator we're going to take one more question.

Thank you.

Our next question comes from the line of John Barton from Wachovia Securities.

Please proceed with your question.

Just trying to piece together some of the data on the flash front from comments you've made.

I think I heard you say that the lower densities were tight.

I know at one point you mentioned that the longest lead times were in the 13 week range but then you made another statement that the older technologies are very competitive on a pricing perspective.

I guess I'm putting words in your mouth by making the correlation between older technologies and lower density but if indeed that's the case, it seems to run against the tightness in the market.

If you can just provide us some input there I'd appreciate it.

Thanks.

The big difference between the high end and the low end of the flash market is who's playing.

At the high end of the market the technology is not limited to the players to really two to three competitors.

At the lower end, as Hector pointed out there's still a lot of people in the survival mode and that sometimes drives, you know, pricing behavior.

That are not in line with your logical conclusion of tightness equals better pricing.

Great.

Okay.

Great.

Thank you very much.

We're going to wrap up the call we want to thank you all for your participation this quarter.

Thank you.

Ladies and gentlemen, that does conclude the conference call for today.

We thank you for participating and ask that you please disconnect your lines.