超微半導體 (AMD) 2002 Q2 法說會逐字稿

Ladies and gentlemen, thank you for standing by.

Welcome to the AMD second quarter earnings conference call.

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

Afterwards, you will be invited to participate in the 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 is being recorded Wednesday, July 17th.

I would now like to turn the conference over to Mr. Michael [Heiss].

Please go ahead, sir

Thank you.

Good afternoon, everyone.

The format of today's call will include prepared comments followed by Q and A. The participants of today's call are Hector De Ruiz, our president and CEO, Rob Herb, EVP of sales and marketing, Bob Rivet, SVP and CFO.

This conference call is a live broadcast and will be replaced on the internet at two sites, street events.com, and of course our own site, AMD.com.

A taped phone replay number for North America is 800-633-8284.

Outside of the United States, 858-812-6440.

The access code for both is 20688141.

Before we begin the conference 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 projects and our manufacturing processes are complex.

Current economic and industry conditions continue to make it especially difficult to forecast product demand.

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 the risk factors in our business.

You will find detailed discussions on our most recent 10-Q filing with the SEC.

For your planning purposes, let me provide you with a couple of key dates for future AMD events.

Our third-quarter earnings are scheduled for October 16th, and our AMD analysts' day at our Silicon Valley headquarters is scheduled for November 7th.

At this point, we're going to begin the call and I'm going to turn it over to Bob Rivet, our CFO.

Thanks, Mike.

As detailed in our press release, issued earlier this afternoon, second-quarter sales were $600 million.

This was down 39% compared to the same period last year, and down 33% compared to the first quarter.

The net loss for the second quarter was $185 million, or a net loss of 54 cents per share.

While the second quarter is typically the seasonally weakest quarter of the year, the magnitude of the market weakness was greater than expected.

Desktop PC demand, especially in Europe and North America, was slow.

Flash memory product sales improved in the second quarter based on the strength of the high-end mobile phone market as consumers are buying feature-rich phones built with high-density, high-performance flash memory.

In our microprocessor product line, we sold in excess of 6 million units in the quarter.

Microprocessor revenue was $380 million, a decline of 35% compared to $588 million in the first quarter.

As mentioned in our April conference call, the number of PC processors shipped in the first quarter had exceeded the number of PCs sold.

Unfortunately, PC demand in the second quarter did not materialize as expected.

Due to this, we took action to lower PC processor inventories in the PC supply chain to help better position our customers in the current market environment.

This resulted in lower second-quarter unit volumes and lower ASPs.

At this level, we believe we may have given back a small amount of market share with respect to processor shipments in the quarter.

However, we believe we maintained share in terms of consumption of processors in PC systems.

Revenue in the memory product line was $175 million in the second quarter, up 9% compared to the first quarter.

Flash memory product sales improved in the second quarter based on the strength of the high-end mobile phone market, as consumers are buying feature-rich phones built with high-density, high-performance flash memory.

For the third quarter in a row, unit and bit shipments increased sequentially, with bit shipments reaching a record level in the second quarter.

Gross margin on a pro forma basis was 7% for the quarter.

The decline was driven by the steep drop in sales, both in units and ASPs, of our PC processors.

Expense management is on track to our plan.

Total second-quarter expenses were less than in the first quarter.

Our profitability issue is solely related to the top line.

As planned and previously announced, we shut down fabs 14 and 15 in Austin, Texas, and restructured [inaudible] Malaysia on schedule.

The full effect of these cost benefits associated with these actions is now positively affecting the current quarter.

Our current head count is at about 13,700 employees.

This compares to over 15,000 employees a year ago.

Research and development spending was on plan at 178 million for the quarter, up 4% from the prior quarter.

Marketing, general, and administrative expenses was also on plan at $160 million for the quarter, up 2% sequentially.

We continue to manage our expenses and have reduced our break-even point.

It is now below 900 million.

Our balance sheet remains strong, with cash balances in excess of $1 billion.

Capital expenditures were 172 million in the quarter, compared to 199 million in the first quarter.

AMD's inventory was flat compared to the first quarter, and is well positioned for future quarters.

Based on the size of the loss in the current quarter, the tax rate benefit was 40%, and is expected to continue at that rate for the balance of the year.

For your modeling purposes, we anticipate R and D expenses will remain flat to up modestly, compared to the second quarter.

In light of current market conditions, we are reducing our capital expenditure plan by about 5% to around $800 million.

This cut will come from non-fab-related capital reductions.

This, however, does not alter our major initiatives.

Our product and technology road maps are on schedule.

Hammer develop road maps remain on trap.

Our manufacturing facilities are performing at world-class levels.

Our cost control initiatives are continuing, and our balance sheet remains strong.

Now, let me turn it over to Hector.

Thank you, Bob.

You know, this is not exactly what I envisioned as my first earnings call.

However, despite the financial facts which are undeniable, I want to make sure that we take the time to discuss those things that are going to position us well for success going forward.

Clearly, from the discussion that was presented by Bob, the second quarter for the PC industry was what some of you have described as the perfect storm.

A weaker PC market environment than anticipated in the already seasonally slow second quarter, ASP pressures as mixed skew toward the value space and processor inventory in the PC supply chain higher than desired, given the community demand levels.

While we anticipate seasonality in the PC market to drive modest growth in PC processor unit demand for the balance of the year, we will continue the aggressive stance we took in the second quarter to reduce processor inventories in the PC supply chain.

And while this was improve our operational efficiency, just as important it will help our customers better position themselves in the current market environment.

Our operational units continue to perform extremely well.

In our factories, our line yields, sort yields, and cycle times are truly outstanding and at world-class levels.

In addition to reducing processor inventories in the PC supply chain, we are managing our inventories prudently as evidenced by the fact that in a very difficult quarter, we held our own inventory levels flat.

Our technology transitions are on target.

As mentioned earlier, our fab 30 wafer starts are now totally converted to 130 nanometers, and all of our PC processor output in the fourth quarter will be 130 nanometer technology.

All microprocessor wafer starts in 25 have now ceased and fab 25 is totally converted to flash technology.

In addition, the transition to 130 nanometer flash technology is ahead of schedule and on target for production shipments beginning by the end of the year out of both fab 25 and JV 3 in Japan.

With fab 30 now totally converted to 130 nanometers, we now turn our attention to the planning and execution of the next-generation of [inaudible] technology at 90 nanometers.

We expect to produce samples of it technology by the end of the year and begin conversion next summer.

Our plans with UNC are progressing well, with product qualification of AMD [inaudible] expected by the end of the year, and production beginning in the first quarter of next year.

The manufacturing options provided by fab 30 and UNC in 2003 will allow us to very effectively manage mix and rapidly respond to market demand of our [inaudible] and hammer based products.

On the product technology side, we are excited to be shipping for revenue our first [inaudible] products named AU 1000 and AU 1500.

In the second quarter, we announced the introduction of the AU 1100 which provides the industry with the highest performance, lower power, and highest integration capability available for PDA type of applications at the lowest possible cost.

Our design win activity is strong, and we will see customer product announcements based on these products during the second half of this year.

We introduced [inaudible] technology for flash in the second quarter as previously announced.

We began shipping product to customers, and as we speak, we're shipping 64 megabit [inaudible] product into high-end cellular applications.

This is an awesome technology that will become pervasive in numerous applications across market segments that will require high-density and high-performance.

The design is elegant in its simplicity, resulting in fewer process steps and higher fab productivity, which leads to an unmatched cost advantage.

Our flash business unit is aggressively introducing new products and accelerating technology advances during this downturn in the industry.

As a matter of fact, in the second quarter just ended, one-third of our flash business was based on products introduced in the last six months.

Our [inaudible] processor road map continues on target, and we will introduce AMD [inaudible] XP models 2400 and 2600 during the second half of this year.

I would like this opportunity to remind everyone that PC customers are looking for performance, not megahertz.

Our [inaudible] XP road map continues to offer our customers competitive performance in unparalleled value as evidenced by our success in the August issue of PC world magazine, where six of the top seven power systems honored were based on AMD [inaudible] XP processors.

Our hammer program is on target.

We have been sampling customers now for several months, and the feedback on the product performance and system performance have been overwhelmingly positive.

Our infrastructure partners are executing extremely well, and we expect to ship hammer-based [inaudible] product by year end and formally introduce the product in the first quarter.

A strong ecosystem around Hammer is being created that consists of [inaudible] such as Microsoft, the Linux community, chip partners such as in [Vidia], [Via], and SIS, as well as motherboard partners such as asus, MSI, and gigabyte.

This growing Hammer ecosystem will enable our partners to deliver superior solutions that span the desktop, mobile, workstation, and server markets.

I opened at the beginning with a reference to being in a market that some of you in the field have described as the perfect storm.

In difficult times, great companies do what it takes to position themselves for success when market conditions improve.

At AMD, we're doing just that.

It is often overlooked in disclosures such as this that companies are really made up of people.

In our case, great people, who continue to execute on the strategic thrust that has been and will continue to be the key to our success.

Our continued investment in technology and products has AMD well positioned to benefit from the market recovery that inevitably will come.

This is why I am confident that we have put in place an intrinsic earnings power that will be unleashed as the market recovers and our top-line revenue crosses the break-even point.

At this point, let me turn it back over to Mike to initiate the question and answer session.

Operator, let's start the Q and A, please.

Thank you.

Ladies and gentlemen, if you wish to register a question for today's question and answer session, you will need to 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 wish to withdraw your polling request, you may do so by pressing the 1 followed by the 3.

If you are on a speakerphone, please pick up your handset before entering your request.

One moment, please, for the first quarter.

The first question comes from Adam Parker with Sanford Bernstein.

Please go ahead.

Yeah.

Hi.

I'm just trying to understand your comments about share here, in terms of giving back a small amount of share, I believe you said.

Intel's architecture group was down 10% sequentially and they said it was mostly due to units, not mix and I think your revenues were down 44% in the PC processor group, so if your units were flat in mobile, can you help us at all gauge the shortfall in revenues and desktop, how much of the sequential decrease was from units having lower pricing?

This is Hector.

Let me answer part of it, and I'd like Bob to - Rob, I'm sorry, Rob, to comment too on it.

You know, as Bob pointed out in the financial narrative that he did, we made a conscious decision in the second quarter to reduce the inventory that existed in the supply chain, and in doing so, we ended up, of course, what we call shipping product that it was less than perhaps it would have been if we had not done that.

From that perspective, that of product shipping into the marketplace, we believe we probably lost a couple of points of share.

However, the story is somewhat more complex than that, and interesting, because when we look at the consumption of our products into actual products our customers build and sell, it's a difficult story, and I'd like Rob to make some comments there.

Yeah, two things Adam.

The first, to clarify the question, we don't get into a lot of details on mix and so forth, but we did comment that PC processor sales were slightly over 6 million units this quarter and I think if you remember right, they were right around 8 million -

Right.

So that's about 29, 30% decline, or whatever.

The total decline in raw units is somewhat complicated [inaudible] that we did make a deliberate effort to drain some inventories in the PC supply chain that limited some shipments this quarter, but I don't think it's going to hurt us from a positioning standpoint in PCs are consumed and what processors are in them.

When data quests reports their numbers, it's probably being going to be three or four weeks yet we see Q2, we were running about 21 to 22% of total PC consumption based on AMD processor.

I would expect it to be pretty close to flat on that number.

From a shipment standpoint, if you assume Intel shipped and I don't know exactly what they shipped and you may have a better ship than me but they think they burned some inventories, they told me they shipped into the PC world something around 29 or 30 million processors to the PC space, but they also did Xbox.

It would seem we potentially loss, you know, a point to maybe two points in what are actually CPU shipments into the market.

Okay. all right.

That's great.

That's helpful.

Shifting gears real quick, just separately, how much revenue do you guys need - would you have needed this quarter in memory products in order to be profitable?

Or how much operating loss did you have on that 175 million in revenue there?

We're not going to give the granularity on that so at this point in time let's just let it go that at the corporate level, as Bob pointed out, we need to be just shy of 900 to break even.

Do you need revenues - can you talk about what revenue you'd need to break even there.

Well, as I said before, just shy of $900 million a quarter would be a break-even point.

Okay.

No, I meant in flash.

No, we're not going to -

All right.

All right, thanks Hector.

Yeah.

Your next question comes from Scott Randall with Soundview technologies.

Please go ahead, sir.

Great, thank you.

Hector, a clarification.

It sounds like you did a good job of burning inventory out in the field or letting your customers do that.

As you go into the September quarter, where do you think the inventory is at - how well is AMD positioned in inventory in the field, I guess is the question?

Well, you know, the one - the one parameter that we all would like to be able to forecast better and understand better, of course, is what the market size actually is going to be in the third and fourth quarter.

As we mentioned in our discussion, we anticipate modest growth, but it's really a function of that, as Rob pointed out, we believe we'll sustain share into - into consumption - the consumption of our products into, you know, people that make the stuff and sell it.

But, you know, having said that, if we assume that our assumptions are right, the market will experience a modest growth, we intend to continue doing in the third and fourth quarter what we did in the second quarter in reducing the inventory in the supply chain.

Is it safe to say that in Q2, you'd reduce inventory more than be required in Q3.

I think it will be very similar.

Similar.

Okay.

And then again, following up on flash, can you at least talk in terms of the direction of profitability as revenues went up for the division, please?

I don't - can you try to - rephrase your question a little differently?

Sure.

Well, for example, Intel actually saw a nice increase in revenue but actually their operating loss increased as pricing hurt them.

Were you folks seeing similar dynamics or did you have a lower loss based on higher - higher revenue?

We started - you know, we think that - or believe that the first quarter was a low watermark for flash, begun, you know, modestly improving.

The second quarter was up almost 10% based on the first quarter.

We - we believe going forward a similar modest increases are expected throughout the balance of the year, and the loss narrows with each quarter that goes by.

Okay.

Great.

Thank you.

The next question comes from Tim Mahon with CS First Boston.

Please go ahead.

Yeah.

I hate to beat a dead soldier here, but I'm just trying to understand, Hector, once again, this whole - you lowered processor inventories in the PC supply chain in the second quarter to help better position your customers.

Does that mean that you actually took some products back, or you just did not ship into the channel whatsoever?

I'm trying to get a feel for what that actually means, and the logistics involved in that.

Without, you know, giving any granularity into the specifics, we did both.

We took some product back and we also did not ship into the channel as perhaps we had done in other quarters.

So if you - if you took product back, why wouldn't - why wouldn't we see a dramatic increase in inventories?

Well, I think we've done a damn good job of managing the operation to make sure that does not occur.

Okay.

And then the second question is, on your mirror, specifically your 64 megabit device, can you tell us if that's manufactured at fab 25 and on what technology node?

It is a 170 nanometer product that is being manufactured right now in fab 25.

JV 3 in [inaudible] is also manufacturing product, mirror bit, so right now as we speak, both factories are actually ramping up, and in terms of, you know, to give an idea of the excitement and confidence that we have on this stuff is that both factories are actually right now, as we speak, in the thousands of wafer starts per week on mirror bit.

Well, that sounds exciting.

Can you give us an idea when you think mirror bit bit shipments will actually cross over from your standard flash?

I don't know the answer to that, frankly.

Okay.

Thank you very much.

The next question comes from David Wu with Wedbush Morgan.

Please go ahead.

Yes.

I was calling to ask this question about so far the actions you took to lower your break-even point is really limited to deferring some capital expenditure, right?

I was wondering, as we go through this inventory clearing process in Q3, I assume that should be done by the end of Q3.

What are the conditions for the PC market that we need to get AMD into the - into the black?

In other words, at the level of revenues at this point, it's still quite a bit away from the roughly 900 million of break-even, and I was thinking about what conditions do we have to have into '03 before we see black ink out of AMD?

You know, it's a complex question.

I'm going to try to answer it and ask either Bob or Rob to chime in, if they want to add anything to that, but first of all, the actions we've taken to reduce our break-even go far beyond just delaying some capital.

I mean, it's - we started last September recognizing that we had two factories in Austin that we needed to shut down.

We put that in place, and the last few weeks we ended up finishing our production in those facilities.

So they are shut down.

We also recognized that we had built an infrastructure in Asia that was larger than we wanted.

We've restructured that and reduced the capability, the capacity, basically more in line with what we wanted and needed going forward.

As Bob pointed out in his discussion, that resulted just from the people side alone on a - on a significant reduction in employment for AMD worldwide.

Now, the real net flow, though, has been - here's a point that I want to make very clearly is that although we had around 2600 people that were actually involved in just the few actions that I just mentioned, we also have backfill with a lot of talent and engineering that we felt we needed to continue to do the things we want to do.

We acquired the [inaudible] center in Boston that's heavily focused on infrastructure for Hammer.

We acquired [Alchemy] that has given us the opportunity to expand the breadth and offering of our opportunities and we continue to make investments in technology to be more competitive.

Having said all that, we've done all that while still continuing to lower our break-even point.

And we will continue to prudently manage that going forward.

We are cautious on hires that we make.

We're still hiring key people, but very, very cautiously.

But, you know, we don't have any sloppy business we can sell or discard or factories we can shut down anymore.

We - we've got to work harder at the top line to be able to get it to 900 million plus, and although from where we stand today, it may sound like - like, you know, it's far away, we were there just not very long ago.

So we're confident that the market will return.

We - we believe we're in the sweet spot of many of these important segments, and we expect that will occur.

What - let me ask another question.

With the balk of your revenues past the first quarter of '03 be Hammer based [inaudible]?

No, that would not be an if I remember assumption. [inaudible] first production shipments there in Q4.

You'll see a ramp throughout '03.

It won't be in the first quarter where you'll see that crossover certainly.

I see.

Thank you.

The next question comes from Mark Grossman with SG Cowan.

Please go ahead.

Great.

Thank you.

In the press release for the outlook, it says that you expect sequential sales growth in flash memory, and that you expect sequential unit growth in [inaudible].

I'm wondering what you're thinking on pricing for [inaudible], if you think revenue for processors will go up in the third quarter.

I'm going to ask Rob to comment on the pricing environment, but we do expect the combination of all of the things we're going to do that we talked about already, you know, in line with reducing - continuing to reduce the inventory, the modest expectation on unit growth [inaudible] it's our belief that as a corporation, we will experience modest sequential growth in revenue, and let me ask Rob to comment on the pricing.

Yeah.

A couple of things, and I'll try to tie a few of them together since a couple of questions come came up on the concept of what's the inventory out there.

We noted that in April that in Q1 we saw a fairly significant increase in shipments of CPUs over what the actual consumption of PCs was.

We have very diligently gotten involved with our customers to understand their own inventory levels to try to make sure they were well positioned with the freshest materials they could go to market in the most competitive position possible.

We already said that entailed, you know, taking some returns in some places, changing what was shipped in some places, various other activities.

We're going to continue to monitor that.

You know, as we go through the third quarter and the fourth quarter, that clearly has an impact on what the actual shipping ASP is, but we also want to make sure that, you know, our customers have the right amount of inventory so we can be competitive in the marketplace.

The interesting thing on Q2 is I heard Intel say that they felt they drained inventory somehow.

We track our numbers pretty closely, you know, based on the shipments we make.

We're pretty confident in what we drained, but if you look at what we just said, you know, we think - we said we shipped over 6 million processors in the PC marketplace.

If you assume Intel shipped 29 or 30 million, that says there's somewhere between 35 and 36 PC processors shipped.

I don't know what anyone's, you know, early estimates are for PCs sold in Q2.

My concern it may actually be less than the processors shipped once again.

Part of the uncertainty we have right now is making sure as we aggressively and diligently manage that inventory, making sure we're accounting for, you know, what is happening in the end market that uses these PC processors, and that will have an impact on the overall ASPs as we go forward, but, you know, with the combination of what we're expecting in unit shipments and AFPs, as Hector said, we are expecting some modest growth largely as a result of the seasonality in the PC market.

Growth in revenues?

I'm sorry?

Growth in revenues in processors?

Yeah, we're expecting some modest growth, as a result largely of seasonality in the PC market.

Okay.

Thank you.

The next question comes from Noah Freeman with Brookside Capital.

Please go ahead.

Hey, guys.

Two quick questions.

As we're beginning to see perhaps less demand for high-end processors as the software demand is not pulling the hardware demand behind it like it has in the years past, do you think your market share will increase as the market swings toward maybe what might be your power sector?

And I have a follow-up.

Well, first of all, you know, we - we continue to execute on the road map that provides a very competitive landscape for our products, and it's phenomenon value relative to performance.

We believe that we are in the sweet spot that's being consumed into end products in the marketplace.

So as a result of that, we're bullish that we can hold on, as Rob pointed out, to our share of PC - of processors that are being consumed into real products.

Our long-term view is to gain share, and we believe that with the introduction of Hammer in the first quarter, we - we regain a competitive position in the top end where we believe we will be able to again, one more time, offer a broad array and spectrum to every segment of the PC space.

So in the short term, we are - we hold onto our position.

Longer term, we believe we'll continue to gain share.

Rob, would you like to add to that?

Yeah, just a couple of comments, maybe in clarification.

You know, Hector mentioned it in his opening comments, and the latest issue of PC world, you know, AMD actually held six out of the top seven spots in what they call power systems or higher-end systems.

And, you know, in the value space, we actually held four out of eight of the spots.

So I think we're equally well represented across the entire range of the market.

The issue of what the share is going to look like, as Hector said, you know, our plan is to at least maintain and over the long term grow our share.

That means we have to maintain a substantial base in value-based PCs because that's largely what's driving the market today.

I do believe, though, AMD has some opportunities in the result of we continue to make gains relative to the commercial space in the small and medium business world.

I believe we have an opportunity as it continues this year and with the promise of Hammer to capture a design of one of the tier 1 manufacturers in the enterprise space.

We continue to grow our share in the mobile space.

And Hammer is, by far, going to be a superior solution in the server space, and I think it's our first real opportunity to engage in what is the - certainly the most profitable piece of the business, and right now, probably the fastest growing piece of the business, albeit a relatively small piece of overall CPU shipments.

Great.

A follow-up question.

Is that a sector that I've always thought of as very strong for your guys has been Asian white box because your competitors don't have the ability to sort of manipulate high-end server - high-end processor availability to those guys because they don't buy the super high-end processors but then the interesting things that I wanted your thoughts on is that Intel yesterday claimed they were gaining share in that sector which struck me as counterintuitive concerning [inaudible] and I'll take the answer off-line.

Thank you very much. thank you.

[inaudible]

The question was in Asia in particular where we've traditionally have been strong because we offer great products to those people that they cannot get from anybody else, his question has to do with our competitor, apparently, claimed that they had gained share in that segment which seemed counterintuitive to him.

Yeah.

Like I said, it's difficult to speculate too much relative to what PC consumption by region looks like right now.

I can tell you, you know, we do quite well in, as you called it, the white box or the do it yourself market in Asia.

In particular, the largest market opportunity in Asia, which is China.

It's also clear to me that my competitor has a whole bunch of low-priced Celerons that he's flooded into the market and so it's quite possible that at any given time he may have grown some share.

But I think our product offering is more than competitive and I think as this market stabilizes, we will continue to maintain and ultimately grow our share in both the value and performance base, quite frankly.

Great.

Thank you very much.

The next question comes from Dan Niles with Lehman Brothers.

Please go ahead.

Great.

You know, maybe I heard you wrong, but I wanted to get a clarification on one point.

You had said in your April release, or Jerry - I forget who - that you thought that processor shipments exceeded consumption of PCs by 4 to 5 million units.

That was your statement back then.

And then I think in answer to one of the earlier questions, you made the comment that you thought that microprocessor shipments exceeded, you know, PC shipments again in the current quarter.

So are you sort of implying that this quarter that we're going into is actually worse in terms of oversupply relative to the past, or did I mishear what you were, in fact, saying, because it was a little confusing.

No, Dan, it probably is confusing.

It's a little confusing to me.

But you sort of heard it right.

It's clear to me - first of all, let me correct the number.

In April, we did say 4 to 5 million at that time.

When the actual numbers came in for what PC shipments looked like, it looks like in April, for Q1, the processor shipments exceeded PC consumption by something closer to 2 to 3 million units.

Okay.

You know, based on what we know we've done, and what at least I heard Intel say, you know, it points to the fact that that situation would have gotten better.

I guess my statement is, you know, with the preliminary information I have, and what I know about the PC market, I'm concerned that it hasn't gotten a lot better, because if you take a look at CPU shipments, we've told you we shipped just over 6 million.

Again, Intel, I'm assuming, 29 to 30 million processors into the PC space.

That excludes the Xbox stuff.

You know, that says there's something in the neighborhood of, you know, 35 million processors shipped.

I don't know what PC sales for Q2 are going to look like but it wouldn't surprise me if it were less than 35 million, which says, again, there were more processors shipped than PCs consumed.

Now, there's some complicating factors, of course.

Some of the processors shipped go into upgrades, so it's not associated with the new PC sale.

There are several types of systems, mostly workstations and servers, that have multiple processors for systems.

I mean, all that has got to be factored in before you can actually conclude that, but I think it's worth paying very close attention to.

Okay.

Maybe tied into that, I guess back to the earlier point, I mean Intel's units, you know, at least from my estimation, was down about 8% sequentially from, you know, the first to the second quarter, roughly.

You know, yours are down somewhere around, I guess, 20% or so, give or take, given, you know, you didn't provide exact numbers.

If sort of the thought is you go into Q3 and, you know, you see modest growth, Intel is obviously forecasting about 4% growth, you know, it seems as though you've, you know, sort of permanently give up some market share as you're going forward, you know, from that, because Intel is talking about having actually reduced channel inventory of their stuff.

So am I missing something or what's - what's going on here?

Because you shipped about 8 million units, right, in Q1?

That's correct.

A couple comments I would make is, you know, Intel is very specific on just channel inventory.

That may be a statement only with respect to their shipment into distribution versus [inaudible].

We're talking more about the overall PC supply chain.

You know, working closely with our customers to make sure we understand what they're sitting on relative to finished good inventory at the PC level, at the motherboard level, at any other point of their operation, and making sure we think with they're well positioned.

When they don't have product that we don't think is well positioned, we're working with them to deal with that.

Again, we are very focused on trying to maintain our shares relative to PCs consumed and the processors in those PCs.

As we get more data here, that will determine what our shipment rates look like, not just in distribution, but even to our OEM customer base.

And we're going to try to work very closely with our customers to make sure they have the freshest inventory possible to be the most competitive in the market.

We think that's ultimately how our customers win and as always, you know, our customer' success is our success.

That's how we're going to continue.

Okay.

And then maybe moving on to a more pleasant topic here, in terms of Hammer versus Itainum II, you know, can you kind of go through the advantages or disadvantages as you see it, in terms of the two architectures from your standpoint.

Obviously Intel went through it in terms of how they're looking at it, but, you know, there's obviously some issues around 64-bit code and how each of the different processors kind of runs it, so, you know, why don't you give us your take on that.

Yeah.

I think a couple of statements.

First and foremost, it's always about the total cost of ownership and there's no way Intel is going to convince the world that the right thing is going to do is change the entire architecture and the entire installed software base.

So anything else they say to me is probably a secondary effect.

However, I will say this: This quarter is a quarter where you're going to see a lot of information relative to Hammer, what its capability is, and the proof points relative to why we think it's the right direction for the 64-bit world.

And I don't want to preempt a lot of what's going to be some very strong PR for us here as we go forward, but we're going to start to get to tell our story relative to ISV partners, IHV partners and other industry partners as we roll through this quarter in Q4.

And do we really need Hammer to get your AFPs going up again, given that sort of for the rest of this year, you know, you're sort of running up the rest of Athalon and, you know, Hammer really kind of moves you back up again, or is there some other stuff that, you know, may end up helping AFPs, given normally Q3 and Q4 are heavy consumer quarters and that tends to imply lower AFPs?

Yeah. [inaudible] quarters is you also tend to get a little better mix on some of the higher-end products.

I mean consumer is where you get the wide variety of high end and low end product.

That season is typically a good time for what I call the power buyers, the people looking for the best machines out there.

So I think it can help in that regard.

I also believe that, again, as we manage through understanding exactly our shipments versus inventory positions, there's some opportunity for us to see some increase there, and of course as we continue to grow our - the share of our business that goes into mobile and workstation and servers, that gives us some opportunity for some up side as well.

So there are - there are some things that do give us opportunity for up sides besides Hammer but clearly with Hammer, we think it gives us, again, a significant differentiated product that allows us up side as we move into 2003.

All right.

Great.

Thank you.

The next question comes from Hans Mosesmann with Prudential Securities.

Please go ahead.

It's actually Brett Pollack for Hans.

Continuing on the Hammer, you mentioned how it's on track which is great but I was hoping you could delve into third party infrastructure products, i.e., chip sets and the schedule for that, if it's, you know, on track and going to coincide with the Hammer introduction.

First of all, just a brief comparison, because I think it's important, is we believe that the learning that we went through in Athalon, which was the very first introduction AMD brought to market that was, I would say, a totally complete different [inaudible] solution in the processor space, that the learning, the relationships established with infrastructure people were really great and tremendous.

And - but in the beginning, those of you who might remember, when we first introduced Athalon, there were times that we had to hold our breath, hoping that the chip set and the motherboard people were going to be there just when we thought.

That has changed so dramatically.

We now have, as I mentioned in my comments, a number of chip set suppliers who actually already have product that addresses Hammer architecture and are eager to continue to develop, design, and produce product going forward.

The same thing on the rest of the infrastructure.

We have aligned - as I referred to in my comments, a strong ecosystem around the Hammer architecture, where our confidence of being - everything being there at the same time, the convergence of all the right things, is very high, and very bullish, that we will meet the launch schedules that we have for early next year for Hammer.

Great.

Looking forward to it.

Thank you.

The next question comes from Eric Rost with Investec.

Please go ahead.

Thanks, guys.

I have two questions rolled into one.

I want to know what you think your share is of the server business is right now overall and how Hammer is going to affect that over 2003.

And then secondly, what you think your share is of corporate PC sales are right now and how Hammer may affect that in 2003 as well.

Thanks.

Let me answer the server Hammer piece and then I'd like Rob to comment on the commercial piece.

First of all, today our - our participation in the server space with Athalon MP, where we have, you know, a multiprocessor systems into the market, has been growing very - very - at some pretty good healthy numbers.

However, it's from a very low base, so I would say that our share of the server space today is - Rob, I'm going to make the point, it's in the single digits, low single digit share

Yeah, right.

And growing rapidly.

The interaction of Hammer completely enhances that picture even further, because we believe that it's going to be a superior product, superior system, superior solution, and we are bullish and we will continue [inaudible] being able to gain share in the server space.

Rob, on the commercial side?

Yeah.

Commercial is complicated so I'll try to answer the question without going into immense detail, but if you look at a worldwide basis, those numbers reported by Dataquest for the first quarter, for a small business which they define as one to 19 plus, we think we had 27% share of the desktop space, and roughly 23% share of the overall space, in combined desktop and mobile.

If you take a look at the hundred to 499 segment, on a worldwide basis again, we think we had somewhere around 20% of the desktop space and 17% of the overall space.

Looking at larger businesses, from 500 to 999 plus, you know, we look like we're about, you know, somewhere between 9 and 10% of both the desktop and the mobile space.

So that sort of gives you a flavor for the different business segments and where the opportunity for growth is, which clearly in the enterprise or large business space, you know, any progress we make towards continuing to get tier 1 support in commercial products is a benefit for us.

The Compaq line and the mobile space lab very, very beneficial.

Some of the commercial in Japan from [inaudible] in Europe have certainly helped us to continue to grow our share in those spaces.

If I can ask a quick follow-up question.

Is - in force, we're hearing some good things about it.

Is it becoming a driver or is it cannibalizing your existing business?

Some of both.

I mean obviously it becomes a driver for people looking for performance and it is a great product, and I think right now, it's establishing itself as a truly differentiated product in the marketplace and I think that clearly will eat into some of the other players at this point in time.

The other players don't sit still.

I mean the good news about competition is competition drives innovation and it drives, you know, choice for consumers, and that's one of the reasons we're such firm believers in competition throughout the industry.

Thanks, guys.

The next question comes from Evan Dooby with LRL capital.

Please go ahead.

Hi, thanks.

Can you explain how the accounting worked for these product returns that you shipped in Q2 and got back in Q3, I assume?

Yeah.

Let me - let me ask Bob to please comment on that.

Thanks.

Those accounting - those returns actually took place in the quarter.

In some cases, that product got reshipped and the return was fairly simple.

It's a reduction of sales and it goes back into inventory.

In a lot of cases, some of that product got shipped again to some other customer, or it is still in our inventory.

So, I mean, that's kind of the simple answer to the question.

Okay.

So it's a mixture of both.

It just confuses me, given the flatness of inventories and the reduction in revenue.

Unless I'm missing something.

No.

You're not missing anything.

I mean, just, you know, a conservative valuation of the inventory as we go forward and adjusting manufacturing rates and the mix all play into that equation.

Okay.

Maybe you can answer this one.

Given the way you ship microprocessors in the quarter and also took product back, where do you think that inventory was - where was the inventory issue?

Was it OEMs?

Distributors?

Stocking agents?

Any idea there?

Yeah.

We have - I think it's a mixture of all of that.

I mean, we have some - some inventory at some of the OEMs that we actually proactively worked with them to to take it back.

We - the same thing on the traditional distribution as well as resellers and brokers.

Okay.

Is - have you guys ever broken out sales by those three categories?

No.

No.

Okay.

And don't plan to do so, I imagine.

Not for consumption outside, no.

Okay.

All right.

Thank you.

The next question comes from Doug Lee with Banc of America Securities.

Please go ahead.

Hi.

This is [inaudible] for Doug Lee.

A couple of questions.

First, could you comment a little bit about flash pricing, how it was this quarter and what your outlook is going into the September quarter and then I had a quick follow-up.

Well, we've made the comment that the pricing pressures have stabilized, and what that means is that the rate of decline on AFPs in flash has slowed down, but it's still very much in a decline, and we believe as we speak there are more learning curve to the decline, more traditional of the semiconductor industry.

They're low.

You know, at this point in time, they're still competitive pressure on flash is, you know, pretty intense, but there's definitely a slowdown in the rate at which the degradation of pricing has been occurring.

Okay.

Great.

And the second question I had for you was actually on the processor mix.

You indicated that it was skewed a little bit towards value processors this quarter.

Could you actually quantify the breakout of value versus performance processors that you shipped in the second quarter?

Rob?

Yeah.

We - we've avoided giving the specific breakdown.

I can tell you [inaudible] as a percentage of total shipment [inaudible] was larger this quarter than it's been for the past several.

Larger this quarter than the past several.

Yes.

Okay.

Thank you.

The next question comes from Jonathan Joseph with Salomon Smith Barney.

Please go ahead.

Yeah.

Rob, sorry to belabor T just want to clarify.

The guidance for units is up, and I would suppose seasonal is up 5%?

Well, seasonal, if you go back historically, is anywhere from about three or four percent up to 14%.

I think the average over the past number of years has been somewhere in the seven or eight percent range.

I would tell you that I would certainly skew to the lower end of the range than the higher end of the range relative to what's going to happen in the overall market.

Okay.

so, you know, I don't think your number is off based on all that market data.

Okay.

And then that - but that's assuming that there's not too much inventory in the channel, and if there is, then maybe that kind of seasonal pickup would not be possible.

Is that correct?

Yeah.

Basically, I think the way to look at it is if the PC market doesn't see that kind of seasonal pickup, again, we're going to manage very diligently, you know, what it is our shipments look like into all channels, OEM, distribution, et cetera, and that can have an impact.

But if we do see the normal seasonal pickups, that says you're going to see Q3 and 46789, and Q4 will be better than Q3, we have to make sure we're well positioned into a growing market so we'll have to watch all that very closely and manage the business, you know, reasonably well to make sure we stay on top of that.

Okay.

Thanks.

Follow-on.

With regard to desktops, you saw flat units in desktops, which outperformed - or I mean note - mobile.

Your mobile outperformed desktops.

What's the condition in the mobile market?

Is it slowing at all, or holding in there?

And then with regard to, you know, what they call a disk top arbitrage, you know, the desktop processor going into mobile, are you seeing much of that?

What do you think of that phenomenon in the market and how it could impact your competitor?

Actually, it's a great point and I should have brought it up when somebody brought up the ASP discussion before.

First of all, the mobile market overall, I think actually while the mobile market was, quote, growing faster than the desktop market, which now may mean declining slower than the desktop market, I guess, clearly I think the mobile market flattened if not declined in Q2.

We'll eventually see all the numbers come out in the overall market.

Our units shipments were flat but clearly one of the pricing pressures you see in the mobile side are people who in the - you know, the large mobiles, the desktop replacement mobiles, you know, use desktop parts and while their preference would always be to get the mobile part at lower power, they sort of demand that the pricing be more closely aligned with what they can get in the desktop processor.

That puts pressure on AFPs, both for AMD and Intel, relative to the products that are acceptable for use in those systems.

So that clearly is having an impact, in particular, on the consumer side of the world.

I don't think it's as true in the commercial side of the world, but that is a phenomenon that's affecting AFPs for the mobile space.

What percent of your traditional customer base, mobile customer base, are doing that sort of arbitrage?

Boy, that's a tough question.

What percent of them.

I think all of them are looking at it.

I don't think all of them are doing it.

Some markets are more susceptible to it than others.

The Japanese market, for instance, there's a whole lot less of that going on because they have a different phenomenon.

The small form [inaudible] global tends tend to be real mobile products.

When you put a desktop into a mobile, you've got heat problems potentially, right, so you have to - you have to crank down the speed and then also you may not be able to put a battery in there.

I mean it's a - it's not going to - it's not going on to take over the whole mobile market but it does seem to have - to be picking up some share.

Yeah, absolutely true.

I think any reasonable mobile enclosure can probably take no more than 45 watts, so your standard desktop parts typically run 60 to 70 watts.

Not all desktop parts are appropriate.

But as you characterize products to deliver them at 45 watts, much of which has been done for the small form factor market in Japan, those products obviously are also capable of playing in what I'll call desktop - desktop replacement mobiles, with lots of compromises.

I mean, yeah, clearly you got to make some trade-offs in other power consumption of the box, including battery life and various other aspects but there's a segment of the market that looks at mobility is not something that they carry around with them all the time.

They typically want it move it from place to place around the home or the dorm room, et cetera, et cetera, and for that segment.

Of of the market, [inaudible] with acceptable performance that's, again, one of the benefits to competition. [inaudible] is great and some consumers have voting that that's an acceptable solution for their mobility needs.

And I know you don't have a number, but 10s%, 15%, or more than that of your mobile space?

I try to give you - no.

Let me answer a different question.

Our actually mobile CPU shipments actually increased as a percentage of [inaudible] it's gone up for the last four quarters in a row but that doesn't answer your question of what are the pressures associated with desktop processors in the mobile market, and how many of the customers are considering and what percentage of their business it is.

I'd be glad to give you guidance if I thought I could.

I really - I really can't give you a number on that because I just don't know.

Okay.

Thanks, Rob.

The next question comes from Michael McConnell with Pacific crested securities.

Please go ahead.

Yeah, Hi, guys.

I was just curious, with the trends you're seeing, low end in terms of [Duron] versus the high end, do you think that this is an emerging trend in the PC industry or are we looking at just kind of an aberration in the second quarter?

Interesting question.

I'll make a few comments and then Rob add to that.

You know, there's - if you do the Warren Buffett thing and go stand at the door of CompUSA or Best Buy or something like that and watch what people do, there's definitely a come lapse of [inaudible] perhaps we had seen a couple of years ago and people are really going after incredibly high-performance valuable products that are, by the definition of we normally talk to considered to be in the value space.

But these products are really pretty awesome.

You can look - and then you look at the variation of price points within a particular processor performance and, you know, depending on the other features surrounding the processor, which is, you know, the hard drive or the memory or whatever, you got a broad range of [inaudible] but they all seem to be collapsing more towards the lower end.

And I'm referring only to the consumer that goes into Best Buy to buy a product.

Rob?

Yeah.

I think a couple comments I would make.

You know, the whole [inaudible] value performance is starting to get skewed especially relative to [inaudible] processor.

I mean people really do care about what is the performance they get for their money.

And when you look at a PC, there are many PC systems with higher performance processors that sell at other perform price [inaudible] in a rich feature set around it.

The phenomenon at the system level, though, has been that the average price for PC systems, you know, has been declining and I believe still is declining.

So I do think we're seeing what is a trend and a lot of it has to do with one of the questions earlier, someone commented that what's the demand driver for needing more performance and, you know, we've had different things over the history of the industry, and right now, we're in that period of time where with what most people do with their PCs, they can do it with a very aggressively priced value PC.

Okay.

Just moving over onto your expenses, in terms of the comments concerning fab 14 and fab 15, you'd said with the shutdowns, you're already starting to see some benefit this quarter.

I was just wondering if you could provide some quantification or where you see in terms of what that could do really to your expenses.

This is Bob.

I'll go back to what we said publicly a little less than a year ago, which was in the third quarter of last year.

By the time we got this closure completed, we would see a 30 million per quarter pickup or reduction in our break-even point, due to expenses.

The fair statement is we've seen some of that as we've gone through time.

We've done it a little bit incrementally in each quarter, but we will see the full impact of that 30 million in the third quarter, though it's not a delta of 30 million from second to third.

Okay.

Okay.

Thanks very much.

And that reduction, just to remind you, is a combination of shutting the facility down, so write-off of assets, and also reduction of people, both in Austin, Texas, and [inaudible] Malaysia.

Okay.

Great.

Thank you.

Operator, we're going to take one more question and then we'll have closing comments by Hector.

Our final question comes from Ben Lynch with Deutsche Banc.

Please go ahead.

Hello.

This is Ben Lynch.

Can I come back, please, to the discussion we had quite a while ago about inventories in the channel and your Q on Q performance and share, et cetera?

Q1 results, you said that the channel over-shipment in the quarter was about 5 to 10% of total.

Your desktop revenues in Q2 were probably down between 45 and 50%, depending on how much mobile was in your revenues in Q1.

So unless you disproportionately over-shipped into the channel in Q1, versus Intel, it's hard to work out why isn't primarily market share gain, not just inventory related but actually end product.

It's really hard to work out.

That doesn't all stack up for me.

And on top of that, just going forward, you are, you know, thinking about flattish revenues in microprocessor or around in there in Q3 but you have spoken about the same amount of inventory reduction in Q3 as you have in Q2.

Okay.

Two comments I would make.

I'm not going to try to dispute any of your numbers because I'm not sitting here with a calculator figuring them out.

I will say this: One way that you can get the phenomenon we currently have is it's yet to be seen, when we see what the PC consumption rates look like versus the processors shipped in, who did the better job of managing the inventory situation.

Right now, we believe we did a very good job of it, and I think it will show when the time comes.

That is why you would see a disparate in a given period of time relative to the revenue percentages that you quoted.

As far as going forward, you know, clearly, you know, we need to continue to monitor that and make sure we're doing the right thing.

All that may be happening is quite frankly, Intel's got a got a larger buildup of PC OEMs.

Now, forget their shipment relative to just their distribution channel.

That's sitting out there in the market and quite frankly becomes, you know, a little bit like stale fish over time and we're trying to do a good job job of working with our customers to make sure that their best positioned with the freshest inventory to be competitive in the marketplace.

But if your over-shipment was in line with the markets in Q1 of 5 to 10%, and let's say you took all of that back in Q2, which was not actually the case, you said, there's still more to go in Q3, that still doesn't get you anywhere near the 44% decline you had in microprocessor revenues coming from inventories.

Yeah.

Well, we also didn't make the statement that all the revenue hit was the unit result of units so you have a lot of factors that you're trying to weigh into one set of numbers there.

Okay.

And then the second question I had was just coming back to sort of balance sheet, your cash amount was down about 200 million in the quarter, also receivables was down around 160 million, your short-term debt was up about a hundred million.

So all that's looking, you know, pretty bad.

There's a net change of about 460 million there in your net cash position.

How comfortable are you going forward for the next few quarters, which are probably also going to be cash draining at an operating level?

How comfortable are you that everything's going to be okay there?

First of all, going forward, you know, in this environment that some of you have described as a perfect storm, we are going to be putting a lot of emphasis to really manage our balance sheet very, very, very judiciously and carefully, and I'd like Bob to make some comments as to the quality of some of the numbers and what you've seen in your observations and put them in the proper context going forward.

Okay.

Thanks, Hector.

Yes, your observations are, indeed, correct.

I mean, I think they're factual of what cash changed by, what receivables changed by, et cetera.

We have a real strong balance sheet.

Obviously, early on in the year, we didn't know what kind of storm we were in, and we did do the convert issue to up the cash balance, $500 million in rough terms, and so we're managing it real prudently, working our way through that.

We did pay off some Dresden debt in the current quarter.

What you're seeing as more of it comes on the book is there's a fourth quarter payment that will come up, but we feel very confident.

We have plenty of cash to work through this, and, you know, maintain all the key programs and initiatives in both products and technology on a go-forward basis.

Okay.

Thank you.

All right.

Well, I just want to take the opportunity to thank you for participating in the call.

We are committed to the product and technology developments that we talked about today.

We're excited about where we are today, and we have never been in a strong position from a product technology point of view as we are going into the third quarter.

We are obviously concerned, as everybody is, as to the difficulty and even able to read and anticipate the changes in the market going forward, but we are - will continue to be prudent in our costs, expenses, and look forward to seeing you in the next call.

Thank you.

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

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