美光科技 (MU) 2004 Q2 法說會逐字稿

Good afternoon, ladies and gentlemen.

Welcome to your Micron Technology second-quarter 2004 earnings conference call.

At this time all parties have been placed on a listen-only mode and the floor will be open for questions following the presentation.

It is now my pleasure to turn the floor over to your host, Mr. Kipp Bedard.

Sir, the floor is yours.

Thank you very much.

I would like to welcome everyone to Micron Technology second-quarter fiscal year 2004 financial release conference call.

On the call today are Steve Appleton, Chairman, CEO and President;

Wilbur Stover, Vice President Finance and Chief Financial Officer; and Mike Sadler, Vice President of Worldwide Sales.

This conference call, including audio and slides is also available on Micron's home page on the Internet at www.micron.com.

If you have not had an opportunity to review the second-quarter fiscal year 2004 financial press release, it is available on our web site at www.micron.com.

Our call will be approximately 60 minutes in length.

There will be a taped audio replay of this call available later this evening at 5:30 p.m.

Mountain Standard time.

You may reach that by dialing 973-341-3080, confirmation code of 4573152.

This replay will run through Wednesday, March 31, 2004 at 5:30 p.m.

Mountain Standard time.

A webcast replay will be available on the Company's web until April 5, 2005.

We encourage you to monitor our web site at www.micron.com throughout the quarter for the most current information on the Company, including information on the various financial conferences that we will be attending.

During the course of the call, we may make projections or other forward-looking statements regarding future events or the future financial performance of the Company and the industry.

We wish to caution you that such statements are predictions and that actual events or results may differ materially.

We refer to you the documents the Company files on a consolidated basis from time to time with the Securities and Exchange Commission, specifically the Company's most recent form 10-K and form 10-Q.

These documents contain and identify important factors that could cause the actual results for the Company on a consolidated basis to differ materially from those contained in our projections or forward-looking statements.

These certain factors can be found on the Company's web site.

Although we believe that the expectations reflected in the forward-looking statements are reasonable, we cannot guarantee future results, levels of activity, performance or achievements.

We are under no duty to update any of the forward-looking statements after the date of the presentation to conform these statements to actual results.

With that, I would like to turn the call over to Mr. Bill Stover.

Thanks, Kipp.

Our second quarter, which ended March 4, was a typical 13-week quarter as compared to the 14 weeks in the immediately preceding quarter.

Second quarter, net sales totaled $991 million compared to $1.1 billion in the first quarter and the Company recorded a $28 million or 4 cent per share loss.

Net sales for Q2 increased 26% over the same quarter of fiscal 2003.

The first calendar quarter of most years reflects softer seasonal market conditions and generally declining average selling prices.

Pricing early in 2004 has held up noticeably better.

The comparison of average selling prices, dependent upon our product mix in our second fiscal quarter, as comparing the second fiscal quarter to the immediately preceding quarter, reflected virtually no change in average selling price.

Gross margin for the quarter came in at 25%, essentially flat with the preceding quarter.

Gross margin from sales of products manufactured in our wholly-owned plants improved during the period, while margins on products purchased under our supply arrangement with Tech Semiconductor had an offsetting decline due to the quarter lag pricing and product mix.

Selling, general and administrative expenses were up slightly from the immediately preceding quarter at $82 million.

Cost containment measures put in place a year ago continues to benefit the Company and SG&A for the third quarter is expected to approximate recent quarterly levels.

Research and development expenses vary significantly with the number of wafers dedicated to new device development and qualification.

Second-quarter R&D totaled $188 million, which was higher than previously anticipated, primarily due to ongoing development at our 300mm line in Manassas, Virginia.

In addition to historical efforts in R&D towards next-generation densities and emerging technologies like DDR2, the current quarter also reflects a higher level of R&D expenses required for diversification into CMOS imaging and Flash product areas.

Our next slide, cash flow slide, you can see cash flow provided by operations remained at a healthy level in the second quarter.

If recent pricing trends are sustained, we currently anticipate noticeable cash flow improvement in the next several quarters.

As of quarter end, Micron had cash and investment balance at just over $1.3 billion and our debt-to-equity ratio remained below 20%.

Capital spending for fiscal year 2004 is currently estimated at the upper end of the range we previously described, the $1.3 to $1.6 billion, current point estimate would be approximately $1.5 billion.

I will turn commentary over to Mike Sadler.

Thanks, Bill.

We are continuing to see solid demand from all sectors of the market for our portfolio of memory products and CMOS image centers.

In the consumer electronics and communication infrastructure areas, this is consistent with the trends we have seen for several quarters now.

In the computing market, we experienced a seasonal strength toward the end of calendar 2003 and then somewhat for a slow down in December.

Our expectation called for moderately slow seasonal demand in early 2004 with the entire year shaping up relatively well.

What we are actually seeing, in fact, in the first three months of the year, I would characterize as much stronger than moderate demand.

We entered fiscal Q2 assuming that computing demand would be modest in the period and have a tolerance to accumulate some mainstream DDR DRAM inventory.

To the contrary, however, demand strengthened as we moved through the quarter and just as was the case in fiscal Q1, we cleared the shelves of finished goods inventory.

As expected, memory concept per system continued to accelerate with a particularly strong post-Christmas snap-back into consumer desktop segment.

Commercial system content growth is steady and also as expected, the gap between content and consumer and commercial systems is continuing to gradually close.

Even with this content growth, the relatively low DRAM prices have left the memory budget per system at a low level of less than 5% of total material cost.

Currently, our customers' demand is exceeding our supply capability and this is creating a near-term upward biased on DRAM pricing.

I have addressed our DDR2 readiness in past earnings calls and analyst conferences and I would like to provide a brief update on this topic today.

Since the last refresh, we have added more component and module alternatives to the portfolio.

We now have over 50 unbuffered, registered and small outline modular configurations, ranging in density from 128 megabytes to 4 gigabytes available for our customers.

These modules are assembled with our 256 megabit, 512 megabit and 1 gigabit DDR2 components.

Our expectations for the demand ramp of DDR2 in the market place are a bit more modest today than was the case one quarter ago.

Regardless of how this market ramp actually materializes, we are well-positioned to provide both volume and product breadth support of DDR2 material for our customers as we move through calendar 2004 and into calendar 2005.

Indications markets are continuing to become a more significant part of our business.

This is attributed to growth of the end markets, and perhaps more importantly, to execution of Micron's product portfolio maintenance and expansion plans.

The market for synchronous DRAM products of all densities is quite robust and all indications are that the markets for these devices, primarily in communications and consumer applications, will have life for quite some time in the future.

We are producing all synchronous DRAM products on advanced manufacturing processes and maintain the ability to support these markets indefinitely in virtually unlimited volumes.

On the new product front, our family of CellularRAM, commonly referred to as Pseudo Static RAM products, is being ramped in production and is well received by the mobile phone customer base.

We are providing this CellularRAM to customers in both discreet form and in multichip packages with our NOR-Flash line of products.

In fiscal Q2, we had another quarter of solid unit and revenue growth from our family of CMOS image sensors.

Micron sensor products are receiving rave reviews from the customer base on image quality and other performance metrics.

The mobile phone and digital camera markets are the primary drivers of demand for our sensors today.

We currently have various VGA 1 megapixel, 2 megapixel and 3 megapixels sensors in commercial production for these markets and are realizing revenues from these products in the current quarter.

We are quite proud of our accomplishments as we have broadened the product offering and extend our position as a key semiconductor supplier in the networking, communications and consumer electronics markets.

Our progress in these markets compliments the solid standing that Micron has established over the years and continues to hold in the computer arena with our mainstream DRAM products.

We are obviously quite pleased with the strength we are seeing in these end markets and the implications that this has for our success in the near to intermediate term.

We appreciate your continued interest and support.

And with that, I will turn it back over to Kipp.

Thank you, Mike.

What we would like to do now is take questions from callers.

And just a reminder, if you are using a speakerphone, please pick up the handset when asking questions so we can hear you clearly.

With that, JT, we would like to open up the line.

Thank you.

Ladies and gentlemen, the floor is open for questions.

If you do have a question at this time, please press the numbers one followed by four on your telephone key pad.

Questions will be taken in the order they are received, and we do ask that while posing your question, you pick up your handset to provide optimal sound quality.

Once again, ladies and gentlemen, if you do have a question, please press the numbers one followed by four on your telephone key pad.

Please hold while I poll for questions.

Thank you.

Our first question is coming from Hans Mosesmann from Schwab Soundview.

Your line is live.

Thanks.

I missed the early part of the call, so if you went through this, I apologize.

Can you give us the updated split expectation for wafer starts by the end of the year for Flash and sensors?

Thanks.

Let me just give you -- this is Steve.

Let me give you a quick outline and we had covered this to some degree at the conference that we had earlier in the year and I think it is playing out as we had said it.

Essentially in the quarter we just ended, we were just over 10%.

We think in the quarter that we are just starting, that it is going to be somewhere around 15%.

And we had indicated earlier that we thought it was going to migrate to 20, 25% by the time we got to the end of the calendar year and that is on track to happen.

Okay.

Thank you.

Thank you.

Our next question is coming from Matt Gabel from Calypso Capital.

Your line is live.

Hi.

Thanks.

I was just wondering if you could comment on what costs did sequentially.

And what shipments were sequentially?

You bet, Matt.

We were roughly flat with costs, and mostly due to a preferred product mix that gave us greater margins.

So roughly flat for the quarter.

Okay, and bit shipped were up 5?

On a weekly comparative basis, yes.

And just slightly up on a quarter to quarter.

You may remember we had a 14-week production quarter in Q1 going to three.

And I think Bill also mentioned, getting back to your cost comment, he mentioned that cost for Micron-owned production went down, and the contribution margin from the tech -- excuse me, the cost went down for Micron.

Cost for product purchased from Tech actually went up a little bit.

So we had two different margin moves there.

One more follow-up.

What percent bit production and wafer starts were .13?

You mean .11?

.11, sorry.

Yeah, .11.

We're about 70% for the Micron production, I think we are probably close to peaking, because we obviously have a nanometer transition as we go through later this year and then we have been allocating wafer starts to things like CMOS imagers that really have nothing to do with the .11 drive on DRAM.

So we feel great by the way, about what we have achieved on that and we're pretty much doing exactly what we said we were going to do.

Okay.

Great, thank you.

Thank you.

Our next question is coming from Joe Osha from Merrill Lynch.

Your line is live.

Hi, guys.

Hey, Joe.

Just in looking at the math here.

Kipp, you just indicated that even leaving the kind of 13 versus 14-week issue aside, that bit shipments were roughly flat sequentially as was price; is that correct?

Actually, bits produced were up just slightly, uh huh.

I believe you said the bit shipments were as well?

I am trying to nail that down because the revenue was obviously down.

On purely leaving the week-to-week comparison issue aside, what did bit shipments do sequentially?

Yes, they have to be down, Joe.

Right, that's --

In Q1, shipments were ahead of production in Q1, so you are off the higher base.

Sure, fair enough.

The second question I have is just on -- if we can drill down a bit more of the current quarter percentage of product that was not DRAM.

I know what the target is late in the year.

Can you tell us what it was for the quarter that just ended?

Yeah, we actually more than doubled our wafer starts towards nonDRAM-related products.

If you go back to Q1, we're talking about 5 to 6% started in nonDRAM products and this quarter, as Steve mentioned, was around 12%, a little bit more than 12%.

That starts obviously -- one presumes that a lot of those products are kind of R&D levels and not necessarily generating revenue yet.

I don't think that is true at all, Joe.

The starts that Kipp is speaking of absolutely are going to generate revenue.

They are not R&D starts.

Okay.

The last question before I go away.

The higher R&D, as you pointed out, that's often times a function of the timing of transition of transition of wafers from sort of qualification to actual production.

Was there a delay in the timing of that for the 300mm fab?

Is that maybe running a little behind where you expect, which is what's dragging that R&D number up some?

Well, as you know, we earlier said that the 300mm was pushed a little bit because we are pioneering copper, so to speak.

But when you think about the total R&D budget, if I get at the question you are really asking, we have been saying now for a quarter or two it's going to come down and it really hasn't.

If you think about what's going on, we have an R&D budget of about $185 million, and we've kind of been running there now for a quarter or two.

About $75 million of that is attributable to 300mm and nonDRAM.

So if you think of the scale of what we have got going on in the diversification and product development or technology development on 300mm, it is pretty significant.

Now, obviously we are in somewhat of a transition moving the wafer starts over to the new products, but essentially, that's where you are seeing the difference between what I think we were speaking about earlier and what's really happening to us.

The 300mm is certainly a part of it but that doesn't give you the total picture.

It's really in combination, those three things are about $75 million for the quarter.

Those were the items that were higher than you expected them to be relative to the margin?

Well, because of the ramp being, I think, actually a little steeper than we thought it was going to be and because of the qualification of more products, if you will, both in image and in Flash, we are just running a little bit higher R&D expenses right now.

We still ultimately expect that to come back down though.

Okay.

Thank you very much.

Thank you.

Our next question is coming from John Lau from Banc of America.

Your line is live.

Yes, hi, thanks.

Mike, you had mentioned that the demand started to accelerate at the end of the quarter and that you finished the quarter pretty bare in the cupboards for the finished goods side.

From a memory perspective, I was wondering, can you comment on the affect of the seasonality currently on the way and notebooks seemed to have been worst than expected.

How was desktops and based upon your comments from the end of the quarter, are we recovering from that notebook inventory issue?

Thank you.

Mike, would you like to take that?

I can take it, sure.

Our best view of notebook demand would be, of course on the demand for small outline dims, the small fore factor memory modules and it's picked back up in the last couple of weeks, John.

Honestly speaking, probably the most drink that we are seeing is in the unbuffered modules for desktops, both consumer and commercial desktops at 256 megabytes and 512 megabyte density.

That has been pretty solid really since the first part of February and it continues to be quite strong.

I mean, that's where we are being stressed the most with respect to our customers pushing us for increased supply, but the direct answer to your question is that demand for small outline dims which should be a pretty good indicator on notebook demand has also picked up quite a bit particularly in the last two or three weeks.

Okay.

Great.

I may have missed it on your call, but did you comment on the bit production goals for the next quarter?

It looks like we are going to start the guidance in the high single-digit range.

And, again, keep in mind, we will still be transitioning more wafers into nonDRAM-related products.

Thank you very much

Thank you, our next question is coming from Tom Thornhill from UBS.

Your line is live.

Thank you.

Actually asked and answered.

Thanks.

Thank you.

Our next question is coming from Michael Masdea from First Boston.

Your line is live.

Thanks.

First question is on price.

Last fall when DRAM got above the kind of $5 level, OEMs obviously produced their bits for box a bit and price fell.

Is that still the ceiling and maybe you can kind of wrap it into your comments earlier about percentage of [inaudible]?

Or do you think we can go higher than that in the current period we are in?

Michael, this is Mike speaking.

I think one of the concerns last fall when prices started to go up was not necessarily -- at least my view, not the absolute price but the uncertainty over where the price ceiling might be.

And I think when price started moving pretty dramatically, customers started to get nervous about that with respect to where things might be headed in the near to immediate term.

This time actually we haven't seen any huge spikes in price, set aside the spot market activity in the last couple of days, but the last two or three price negotiation periods with our OEMs we have been able to lift prices somewhere in the range of 5 to 10% each time.

This is a little bit more comfortable and I think customers take a little more comfort in a pretty modest price moves and I think they are able to digest them better at this point in time than they were last year.

So I don't really believe that the absolute price is the problem.

I think it is the relative uncertainty about where prices might go in the future.

To be perfectly honest with you, our customers don't seem to be too uncomfortable or too concerned about the prices ticking up here in the last couple of months.

Okay, that's helpful.

On the manufacturing side .11, you guys seem pretty comfortable.

Clearly there's been some news about your competitors over the last quarter having problems.

Has that impacted the supply demand balance materials do you think?

And is there any risk that supply can come on quicker than their growth can come on -- excuse me, or the wafer routes and kind of increase because of that?

I can try to address the .11 and Mike, maybe you can comment a little on it.

I think that there is no question that in contrast to what some of our competitors said they were going to do on the 110 nanometer, it just failed to materialize and, of course, that has had an impact on their bit output production.

But I think that as they recover from that, it doesn't happen overnight.

It takes time to get the facilities converted.

And if you look at what Micron is able to achieve even on a relatively accelerated pace, essentially, it took us six to eight months over time to even reach the levels that we did as we started that transition.

So there will be bit growth from it, but I don't think it is something that is so dramatic that happens overnight.

It takes time for that stuff to happen in order to get those additional bits out.

So it's hard for me to comment on what supply others may bring into the market, but when you look at the strength of the demand today, exceeding what appears to be the supply from all the competitors, and some kind of overnight change just doesn't seem that likely.

All right, I guess just the last question to that point is, just what fundamentally and maybe Mike can get at this too.

What fundamentally is driving the better demand picture for the second quarter on a calendar basis just because you guys are increasing your bits most of the time, [inaudible] increasing their bits a little bit and yet we have a PC quarter that is seasonally weak and can you just help us kind of, what gives you guys comfort that we will still be in a state of kind of undersupply in the second quarter?

A couple of comments, Michael.

First of all, if I aggregate the customers demand statements to us or demand forecast us to, in calendar Q2 versus calendar Q1, so sequential growth, they are talking about, again this is ballpark, about 10% bit consumption growth calendar Q2 versus calendar Q1.

And then going forward, calendar Q3 versus calendar Q2, it's more like 15 to 20%.

So those are pretty healthy growth figures.

I think the primary catalyst is content per system, this is in the PC area.

Content per system, as I mentioned we saw pretty healthy snapback in content per system in the consumer desktops.

As a matter of fact, the data that we got, I think the [inaudible] was the source, indicated about 15% content per system growth in the consumer desktops in calendar Q1 versus calendar Q4.

So that's pretty healthy growth and that is driving a large portion of this demand growth.

Also, we have got really continued strength, surprisingly strong demand in the nonPC applications.

SDRAM for us, about 25% our bit consumption or bit production is occurring in SDRAM and obviously that is selling at a nice premium per bit as well.

So the nonPC applications are continuing to be very strong and there's really not much seasonality to the nonPC applications in terms of demand growth.

And I guess, just adding to what Steve said, we are seeing strong demand, really, on all applications.

We are seeing strong demand, but I would not necessarily characterize it as explosive.

So we are always dealing with a supply demand balance or the relative demand supply balance and the conclusion that we would have to draw is that the supply growth is just not really there at least in the intermediate term.

So you guys are not looking for anything special on the demand side, seasonal slightly better but there's a bit per box kind of impact for you guys and mix impact in kind of a niche market growing impact for you guys that's making you look a lot better than what you see normally.

Is that fair?

Yeah, I think that's -- I mean, I think there is obviously a few things at work here.

One of them is that there is an allocation of wafers at least for us and I think there are for other companies like Samsung where they are trying to meet demand and applications across a customer base that is much broader than just the PC market.

So that's impacting it.

And add into that some of the growth that Mike has spoken about which was beyond maybe what we have seen historically the last couple of years.

And it makes for the kind of environment that we have right now.

All right, thanks.

Thank you.

Our next question is coming from Nimal Vallipuram from DRKW.

Your line is live.

A couple of questions.

First of all, just going back to those two questions that Joe asked earlier.

If you can just give us some more idea.

Number one, there seems to be a trend among the DRAM vendors including Micron to switch from [inaudible] to other markets which are more profitable or for strategic reasons like Flash market.

Is there any internal target as to what percentage -- to what percentage can you go for DRAM or are you willing to be more opportunistic as to where to put the wafers going forward.

You bet, Nimal, that is a great question.

In fact, as outlined in the quarter, we just finished, Mike was able to choose a product mix that kept ASPs virtually flat when I am sure you realized December, January, and early part of February was a downward trend in pricing.

So he has been able to modify the production moves into products that provide better margin for us, and so that is something that we can continue to do going forward.

I think that answered part of your question.

Why don't you repeat the second part.

I mean, is there any kind of goal as to whether you will go up to 70% DRAM, 30% Flash and CMOS or anything like that or more opportunistic as you see what the market is going to do?

This is Steve.

I think we have to differentiate between -- when you say "opportunistic," one of the challenges we have and I think any of the other large DRAM producers is, we have a customer base that we have committed to support, and so if you think of the computer environment historically being the dominant customer base for this product, we will not simply be opportunistic and not support them in order to go out and get some kind of higher margin in a shorter-term period knowing that, of course, we have long-term ties with them and we have to continue to support their business.

So when we talk about moving wafer allocation around and having other products come on line, it really is a gradual transition for us.

It's not an overnight transition, and we have commitments to meet with our customer base.

Our approach is not so much opportunistic as it is a longer-term repositioning of the company in terms of its product mix.

So we don't have a particular target.

We need to see where the relative demands for the application settle out in the next year or two, and that will determine where we think we should be.

Just on the second question.

If you look at your 110 nanometer wafers, you said like 70% of that and you might be peaking right now.

Can you give us some sort of benchmark as to what it is going to be by the end of the year in terms of 90 nanometer wafers, as well as what is the plan for the 300mm output.

Well, the 300mm output we have been pretty consistent.

Our first goal is to ramp it 10,000 to 15,000 wafers per month that will get to that point sometime in, I would say the first or second quarter of '05.

And that's on track.

We have equipment that's being installed between now and then, between now and the end of the year in order to make that happen.

I am sorry, what was the other question?

As far as the 90 nanometer wafers are concerned, are there any benchmark as to what it is going to be by the end of the calendar year?

Oh, yeah, sorry.

We are not really giving much data on that in terms of where we are going to be at the end of the year because obviously we haven't started that transition yet and the technology's still being tuned up, if you will.

So I think as we get towards mid-part of this year, we will be ready to give you a better indication of what that is going to be.

Obviously we will start the transition as we go towards the end of the year, but I don't have good data for you right now.

Just finally, according to your intelligence of what you are doing on the R&D and what you are hearing from the other people.

Given the so-called problems some of the other companies have had in the 110 nanometer yields, going to 90 nanometer, would that create any more problem?

I guess my question is this.

When you move from 130 to 110 to 90, is getting the right yields getting a bit more complicated for the DOM companies.

We don't think it will be for us, but we do think it will be for others, in particular maybe for one of the other technologies.

All right, thanks a lot.

Thank you.

Our next question is going to come from John Barton from Wachovia.

Your line is live.

You made the statement that DDR2 ramp is not happening as fast as you had anticipated last quarter or something along those lines.

Can you elaborate on that, what you think is causing the change and what kind of time frame you are looking at for DDR2, please?

This is Mike speaking.

By the way, the time frame is not really changed significantly.

I think it is the relative slope of the ramp throughout the balance of the year.

We had been expecting perhaps as much as 30% of total demand but going out a year, being in the computing area being for DDR2 and my take now is probably more like 15 to 20% is probably a realistic expectation.

There are a number of issues.

Probably most significantly commercial issues.

I don't think the customer base is prepared to pay significant premium in very high volume for DDR2, and quite frankly, we are of the belief that a significant premium is appropriate for DDR2 relative to DDR1.

There are some pretty significant cost increases on DDR2, not just die side, but also some infrastructure costs in terms of additional test equipment and so forth.

So I think primarily commercial reasons are going to result in a slower ramp but I don't think there is any question whatsoever that this is going to be the mainstream technology in the PC arena as we move through 2005.

Actually not just the PC, the server arena as well.

Bill, if I could, from an inventory perspective, any benefit for previously written-off inventory this quarter?

Inventory grew right around $40 million.

What that was composed of and what you think you will be doing with inventory next quarter please?

It is in the teens of million in the current quarter, so not very significant.

There's $16 million yet to flow through in two quarters.

And three quarters of that will probably come through Q3.

And then inventory-wise, still attempting to build more inventory going forward?

Or what are you thinking there?

You are really speaking to Mike and what the market conditions are.

Our inventory --

Well, I think -- this is Steve.

Let me say we absolutely have to over the next couple of quarters, whenever we decide it is the best time to do it, we can't run at the low levels of inventory we have right now and adequately service the customer in terms of timing on the demand.

Mike can comment on it further, but we will have to put some buffer in there instead of the way that we have been running because it just makes it very tough for us to adequately service the customer.

I think to be a little more specific on that, the last two quarters we have run finished goods inventory levels down to less than one week worth of supply and taking into account today's business models that really require us to keep inventory at our customers manufacturing sites, it is very, very difficult to provide a reasonable level of service with less than a week's worth of inventory.

We think the number should be more like about two weeks worth of inventory.

As Steve mentioned, in order to improve our customers satisfaction levels and provide better service to our customers, we are going to have to kind of rebuild this inventory over the next couple of quarters.

Thank you.

Thank you.

Our next question is coming from Andrew Root from Goldman Sachs.

Your line is live.

I want to explore the ASP issues for a second.

You mentioned some incident in terms of talking about mix and being opportunistic.

I am wondering if there is any contribution as well to lineary in the quarter, if you shift a higher percentage of bits in the last months of the quarter attempting to build inventory early on and basically clearing the shelves at the end of the quarter and relatedly if that had anything to do with DSOs popping up or if you could explore that issue a little bit as well?

I can take part of that and maybe Mike would like to jump in.

Again, if you look at Mike's comments that as we exited the last quarter and again this quarter, we had finished goods inventories at extremely low levels, that creates a nonlinear aspect in the first couple of weeks in the next quarter anyway.

So part of your question is dealt with that way.

I think if you go back to Mike's original comments, he also saw an improving demand scenario as he got through the quarter.

Maybe, Mike, you would like to expand on that.

I would be happy too.

It's interesting.

The first couple weeks of the quarter -- keep in mind the quarter we are talking about is December, January and February, the first few weeks of the quarter were relatively strong as we continued to feed inventory to our customers, primarily I believe to support the end-of-year or Christmas selling season in the PC area anyway.

And the latter part of December, the first part of January were relatively weak with holiday hit and Chinese New Year approaching and so forth.

And then February was a very strong shipment month.

So I am not sure what you would do with that information, but as prices were increasing in the latter part of February, our shipping momentum was pretty strong.

Okay.

Then the DSO angle, just curious?

DSOs were up a little bit.

I am sorry, DSO?

DSOs, days sales outstanding were up a little bit in the quarter.

Nothing unusual was at play.

Okay.

Finally just on the mix, the percent of product that went into the spot market and the percent that went to DDR versus SDRAM.

On the Micron-branded material in the quarter about -- somewhere between 5 and 10% of the material went into the spot market, and a large portion of that was toward the end of the quarter.

Of course, we do put product out into what is commonly referred to as the spot market through our crucial channel and our spec-tech channel as well, that would all be added into that 5 to 10%.

On the technology mix, about 25% of our bit output, give or take a percentage point or two, 25% of our bit output was synchronous DRAM with the balance for all intent of purposes being DDR.

There was a small amount of EDL and a small amount of DDR2 as well but.

Then finally, [inaudible] the top of the market or were, any comments on channel inventories and away from Micron, what you are seeing in terms of available supply out there?

I am pretty confident based on some of the conversations I have had with customers in the last 72 hours or so that there is nothing in the channels.

Nothing, okay, great, thank you.

Thank you.

Our next question is from Ted Parmagiani from Lehman Brothers.

Your line is live.

Yeah, it is Parmagiani.

Thank you.

Good afternoon.

Guys, I just wanted to ask a question that had to go to your costs and improvements that you might be able to obtain here, given the stable-to-better pricing environment that you expect here in the quarter.

And then also, just a follow-up on DDR2.

What impact would that have on your overall pricing and strategy if, in fact, even, say, a mild scenario or a kind of a midpoint of what you are thinking about in terms of mix takes place here in the second half of the year?

I will take the second one first, if that's all right, Kipp?

You bet, Mike.

Go ahead.

Mike speaking.

On the DDR2 ramp, to be honest with you a more conservative ramp of DDR2 in the marketplace through '04 is probably in our favor from, let's say a net income standpoint.

We will put our most cost-effective DDR2 devices, the 6F square [ph] devices into volume production towards the end of the calendar year, ramping really through the first part of 2005.

To the extent that the market really waits and starts to ramp up aggressively when we can support with volume with the 6F square devices, we're going to be in a much better position from a costs per bits strandpoint to achieve some relative benefit there.

We actually prefer a more modest ramp.

Now having said that, no matter how steep the ramp is that the market wants to take on, we are in great position to support it.

We've got -- I think I can confidently say we are positioned better than anybody else in the business right now to support a wide variety of DDR2 module configurations.

Thanks, Mike.

Can you repeat the cost question again?

Yeah, since pricing isn't really going to be -- at least not a negative variable or at least that is a good assumption might be for this quarter, in terms of looking at the quarter from a gross margin standpoint, that pretty much leaves everything to cost in terms of maybe some of the potential upside you may be able to drive in the quarter, and I am just wondering where -- should we expect some improvement in your cost this quarter.

For example, is there a lot of potential in yield improvement that could maybe move margins here incrementally a little bit given the stable environment on the top line there?

I understand.

Let me point out a couple of things we have mentioned so far to at least give you direction and then Bill will hop in for a little more detail on it.

Now we have talked about an improving percentage of .11.

We have yet to mention the 6F squared percentage of starts.

Keep in mind when we are talking starts, somewhere between 50 and 60 days later they are hitting our back end so they are starting to flow through.

So we have just gone from a Q1 period where we had roughly 30% of our starts at .11 to now we are up to almost as mentioned before around 70%.

In December we went up from up to 40% starts on F squared and that's now running 50 to 60%.

So that will give you some idea of the metrics at work that can impact cost of the next couple of quarters.

And then Bill, maybe you would like to put a little more detail around it.

Just layer one other item on top of it, from Mike's commentary, you get an indication of the improving ASP environment, and as we have consistently disclosed, our supply arrangement with the Tech Semiconductor has a lag quarter pricing.

So you can see that in an improving ASP environment you will see some improving margin off of that supply arrangement.

Okay.

Great.

Thank you.

Thank you.

Our next question is coming from David Wong from AG Edwards.

Your line is live.

Thank you very much.

Can you help us understand with your current Capex guidance for expectations what you expect production, bit growth is going to be the fiscal year 2004?

We are still looking in the 35 to 40% range for DRAM product and, of course, within that, we started our fiscal Q1 with about 4, 5, 6% of wafers aimed at nonDRAM product and, as Steve mentioned earlier, that can be as high as in the 20 to 25% range as we move through the year.

Right, thanks.

Thank you.

Our next question is coming from Menish Goyle from Neuberger Berman.

Your line is live.

Two questions.

First, if you are running roughly 10 to 12% [inaudible] in nonDRAM products is it reasonable to expect during the current quarter just about 10% revenue was from nonDRAM products?

Well, there is a little bit of a lag, of course, between start and revenue, so I think the way to look at it is whether it's -- some of that stuff we sell in wafers form and some of it actually makes its way all the way on to an NCP which is beyond kind of the back-end packaging.

So you have to figure somewhere between 60 and 90 days later, it will convert into revenue.

So I think it is a pretty good assumption, I can't give you an exact timing on the lag, but it varies 60 to 90 days.

So if you look at the revenues of your nonDRAM products, how would you classify the profitability of that revenue if you exclude one-time items.

Do you think it is close to your DRAM profitability?

Or is it very different?

Yeah, for the mainstream DRAM, it's better margin.

The operating income line and gross margin line?

Yeah, sure.

We are ready to break it down for you.

Flash in current pricing comparatives, okay.

We have to make sure you understand that.

The NOR-Flash is just slightly less gross margin and we have both the image sensors and the Pseudo SRAM which are better gross margin.

I see.

Thank you.

And one more question, if I look at your cost of goods, and after adjusting it for the inventory charges for the last quarter and for the current quarter, it seems like your cost of goods declined roughly by $80 million.

Is that correct?

And then secondly, what caused that decline?

If your depreciation is roughly flat?

I think we will have to work with you offline, Menish as I don't have those inventory adjustments in front of me.

Seems like -- okay, that's fine.

Thank you.

Depreciation for the period is about -- down about $15 million, just on the 14-to-13-week.

But otherwise we'll have to get those --.

So about $70 million sequential decline in cost of goods.

Maybe we can take it offline.

Thank you.

Thank you.

Our next question is coming from Clark Westmont from Smith Barney.

Your line is live.

I think, Steve, you covered pretty much everything.

Maybe one just last return to the notebook side.

Can you distinguish much between the notebook and desktop demand and what you have seen in notebook in particular in the last, say, three to four or five weeks?

Yeah, I think that question came up earlier.

We have seen some -- unfortunately, I don't have data at my fingertips to specifically reference, but we have seen a pretty reasonable bounce back in demand for small outline dims that would be used to populate notebook computers with memory.

We have seen a pretty strong bounce back in the last two or three weeks.

I wish I could tell you what percentage of our overall output was going on small outline versus unbuffered dims for desktops, I don't have that data, but I am sure we can get it to you offline.

I appreciate it

Thank you.

Our next question is coming from Krishna Shankar from JMP Securities.

Your line is live.

Yes.

Can you give us your best estimate for industry bit production growth demand for this year given your assessment of 300mm [inaudible] among your competitors.

I didn't quite get it all.

I am going to repeat it and you can tell me if this is correct.

Were you looking for an industry supply bit growth?

Yeah, supply bit growth versus demand from your viewpoint.

We've seen some recent estimates by those who follow that and model those things and we are now looking at a range that used to be in terms of bits produced for the industry for '04 over '03.

We started the year looking at expert estimates in the 50% to 60% range and now we've seen them as low as 35 and as high as about 50.

And we've got demand -- you can kind of -- you will have to calculate your own, but I think Mike's comments earlier about Q1 and Q2 give you a pretty good lead-in to what potentially the calendar year could be on the demand side.

Okay.

Can you give us what your 6F square production was for the most recent quarter as a percent of total DRAM.

We have given it to you in terms of starts.

And, again, as Steve mentioned, lag that by about 60 to 90 days.

So if you recall, when we were on our December conference call, September was about 10% starts, and when we got into December, we had raised that to about 40%.

So a pretty steep ramp right at the end of the calendar year end.

Okay.

And final question, with all your 90mm be on 300mm wafers or will you be transitioning some of your 200mm capacity to 90 nanometer technology.

No, it will transition on 200mm as well.

Thank you.

Thank you.

Our next question is coming from Dan Niles with Neuberger Berman.

Your line is live.

Thanks.

Steve, this transition to .11 that you are seeing for the industry, is it any more challenging than some of the other transitions that you have seen over your 20 years in the business?

Because it always seems to be every other node, there seems to be some issue sometimes in terms of --

Well, Dan, I think it depends on your starting point, and for us, the .11, went extremely well and we have already talked about the .13 did not go as well, but we had a lot of focus on .11 earlier on and we made that choice, and as a result, the 90 nanometer for us is going to be relatively straightforward as well because we are not going through a big Litho-change in order to do it.

I think many of our competitors do have to go through a change in order to make that transition.

It also turned out that the .11 was much more difficult for some others than maybe they thought it would be, and if you remember, there were a couple of them out there bragging allowed about .13 and how that was going and what percentage of production they were going to hit on that and if you back up a year ago we were kind of the first company to come out and say what we were going to do on .11 and we gave targets on percentage of starts that we were going to hit throughout the year and, of course, history now shows we hit those.

At least a couple of our competitors also gave targets of what they were going to hit on .11, and they haven't hit them.

In fact one of them is just way, way behind of what they said they were going to do and I think that is because of the difficulty you are mentioning on trying to make the transition from 13 to 11.

.11 to 90 nanometer, frankly, I think is going to be tough, maybe even tougher for some.

We don't believe that it will be any more difficult for us, and I have already mentioned we had don't have to do a litho change to do it.

We feel pretty good at where we are at, but I don't think that is the case for everybody.

Are most of your competitors not having to make a litho change to get the 11 or a double litho change to 11 and then to 90?

At this point, I think most of them won't have to make a litho change as well once they get the 11 ironed out.

All right, I guess a separate question on the bit production so I understand it better.

Kipp, I think you had mentioned bit increases in the high single digits, is that DRAM, Flash, Pseudo SRAM, that's all memory and then should we add the CMOS image sensors on top of that?

What is the right math?

That is correct, Dan.

Okay.

Thanks.

Thank you.

Our next question is coming from Adam Parker with Sanford Bernstein.

Your line is live.

Hey, this is Nathan Dupree for Adam Parker.

A question on Capex.

You had only $468 million for the first two quarters out of the $1.5 billion guidance.

Can you just help me with thinking about how that will hit for the rest of the year, and more interestingly, how that will hit the cash flow statement over the year?

In the numbers that we are giving you, they are a pretty good correlation with the cash flows so you can see there is a significant amount hitting in the second half of the year.

As far as just the composition of that, it is very, very heavy to fabs, back end pretty light, actually, and there's about an equal split between 300mm and the other fabs generally.

Leave it with that kind of a broad characterization, but you are looking at it correctly there is some significant cash flow in the second half of the year.

Okay.

And if that does flow the way you expect and you do get the qualifications at the time you expect, what are you looking at for depreciation then?

We are still giving an estimate for fiscal year '04 at $1.2 billion, running the $300 to $315 million going forward on a quarterly basis as a good estimate.

Okay.

Okay, different question.

About average contract prices.

For March, do you have any comments on the first half and second half of March?

First half on percentage-wise, we were able to raise prices from the second half of February by about 7% or 8%.

And the absolute prices in the low to mid 4s on a 256 megabit equivalent basis and for the second half of March same thing -- I'm sorry, the pricing I just referenced was for the second half of March.

Okay.

For the second half of March able to raise them 7% or 8% versus the first half of March, and we will have another price negotiation with our OEMs next Tuesday, and my expectation is for a price increase there as well.

Okay.

And is -- so the tech JV, just to get the pricing, that's the quarterly in its lag, that's based off of contract pricing, right?

Not the total mix?

In other words, it does include just contract or contract in spot?

It is an average based on the type of devices that are running, which is dominantly DDR devices.

Okay.

And so that should, therefore, be up based on the increase in prices this --.

Almost all of our stuff is going to contract these days anyway so.

Okay, thanks

Thank you.

Our next question is coming from Quinn Bolton from Oppenheimer.

Your line is live.

Hi.

First, a quick clarification on the contract pricing just discussed.

You said it was up 7% to 8% in each of the first half of March and second half of March.

I wanted to get -- where is the pricing now on the contract, is it low to mid-4s or is that as the first half.

I am sorry, I know I answered that in an unclear manner.

It is currently in the low to mid-4s and that is for the second half of March.

Okay.

Great.

And just a clarification on the inventory levels.

It looks like you shipped pretty much everything you built.

We are wondering if you shipped -- squeezed anything more out of the inventory during the quarter.

Whether your inventories were effectively flat quarter over quarter so you just shipment matched production in the quarter.

Shipments matched production.

Okay, thank you.

Thank you.

Our next question is coming from Mora Arabi from Rosetta Group.

Your line is live.

I would like to ask how much of the revenue is coming from crucial and the aftermarket.

We couldn't quite hear, what part of the revenues came from what?

Crucial and aftermarkets in general.

The crucial business unit represents approximately 10% of our revenues and about half of that would be administering an aftermarket upgrade program for our OEM customers and the other half would be literally selling modules one by one or in small volumes to individual consumers or small businesses.

Okay, the follow-up just on the cost question.

You had a target before, I think, to bring costs down 40% or something of that magnitude.

Are you ahead or on track or?

Well, when we say "cost target" you have to back up really one year, and frankly, I can't remember exactly where it is right now, what it was, but we had achieved it.

I think we said we were going to try to hit somewhere to around 40% for the year, and then the question is, what happened on a go-forward basis.

Right.

We have been able to reduce our cost by 25% to 30% per year on a normalized curve and, of course, we're going to continue to try to drive down that curve just like we have in the past.

So this year you don't have the 40% goal?

Well, it's a little deceiving because you have to talk about the product type now.

As Mike Sadler had mentioned on a lower density products even though we will, of course, try to have cost reductions on those as well, you don't have the benefit of going to a higher density because the particular density that we're at, say it's a 64 or 128 meg, that is the density you are at.

You may shrink it but you're not going to a higher density which gives you more bit.

It is a little bit deceiving to talk about what the cost targets are, because it is not apples to apples anymore.

Okay.

Thank you.

Thank you.

Our next question is coming from Ben Lynch from Deutsche Bank.

Your line is live.

Thanks.

Hi there, guys.

I seem to recall on the Q1 conference call you said your Q2 bit production would be sort of flat, not including CMOS image sensors.

And also accounting for the sort of 14-to-13-week change.

Now you are saying that the weekly volume was up about 5%, which, if I just take the one-week difference, it would seem to be a little bit lighter than you had guided to.

Can you help me understand that, please?

Actually we did hit what it was.

We were actually up a couple percent.

I think it was one or two percent on an absolute basis, Ben.

Okay.

5% on a weekly basis.

Okay.

Right.

I had understood that it was even despite the number of weeks being down.

That's correct.

If you compared 13 to 13, you can figure that, a 5% increase per week.

If you did absolute to absolute, we were up just a little bit better than what we thought.

We thought we would be flat and were actually were up 1 to 2%.

Okay.

Just in terms of the cost trends.

A lot of your overall wafer volumes in this quarter were nonmainstream DRAM, [inaudible] Flash, also the higher margin SDRAM parts you are talking about.

If I add up the SDRAM you're saying is 25% of output.

The nonDRAM about 12%.

That gets me to 37% or so.

I know I am mixing up wafer starts and wafer outs, but it's a high portion of your volumes.

Is that right just before I get on with my question?

Actually if you go back to what we guided before, if you basically just leave a one-quarter gap in there and go back to what we talked about, we were 5, 6% starts on nonDRAM last quarter which means that is roughly the nonrevenue number in the quarter we just reported.

So 30% of wafer shipments were nonmainstream DRAM.

Well, you also now have to figure in that SDRAM had higher ASP.

So you are now combining bit ship and revenues.

So you're going to have to go in one direction or the other for me.

Okay.

It is a complicated mix, but I guess where I am ultimately getting at is, you are having a lot of .11 Micron coming on line now and or you are shipping out in this Q3, 6F squared is really starting to ramp as well and yet you are only guiding despite obviously the continued ramp of image sensors, et cetera, production volumes to be up high single digits.

I would have thought it would be up a lot more because you are getting two massive bangs for your buck with the .11 and 6F squared.

Are you at the yields yet to get the maximum benefit from those?

We are pleased with the yields.

It's really two things.

You have, one, additional wafers shifting towards nonDRAM related products, and number two, a couple of component shifts as well.

As Mike mentioned, a little more DDR2 which is slightly less efficient than other things and then in the margin category or the contribution category, some of the trailing edge products like 128 megs or 64 megabit DRAMs may not have the bit impact but certainly has the margin contribution impact.

Okay.

And maybe just my last question.

When you look at all the DRAM guys who report their numbers, other than Infineon who I think you are alluding to as the major guy having problems with .11.

Everyone else is sort of achieving pretty decent operating margins at the moment.

Samsung supposedly 35%.

Some of the guys in Taiwan, 20% or so.

How do you think they are managing to do that, Infineon and you guys are here close to break even.

You nailed it.

We are right in the middle of that.

You say they are shooting margins to around somewhere between 20 and 35%.

I think we turned in a quarter at 25.

I am talking about operating margins.

Not gross margins.

Well, again, if you were going to compare to the Taiwanese, they don't have the SG&A and R&D numbers.

So again, you will have to compare like numbers.

Okay.

Yeah.

They will say fully loaded costs are giving [inaudible].

That is what they are reporting in their accounts, before going to pretax income.

Okay.

I just thought I would try to get your perspective on why at the moment they seem to be a lot more profitable.

Okay.

Thank you.

Thank you.

Our next question is coming from Ty Gwenn from Susquehanna.

Your line is live.

Just one question.

I was wondering with your anticipation of the undersupply market for the second half, I was just wondering, Steve, do you have plans to pull in your 300mm production into the 2004?

Well, we are already on probably as good a pace as we can be on.

If you think back to when we struck an agreement with the Intel investment, we said that it had to do with 300mm investment.

And at that time, we had made a decision to try to bring in the 300mm ramp to the best of our ability, and that's just the schedule we are on.

If you go look at lead times on steppers and those kind of things, we are going to have to live with that.

And, the equipment lead times now are, they've pushed out six to 12 months, maybe longer in some particular cases.

So you have to back up and realize we ordered a lot of that equipment in the fall but we don't get delivery of it until this fall and that is just the schedule we have to live with.

Okay.

And then on your CMOS image sensor, could you elaborate more on where you are today in terms of like design win and penetration in terms of, like, digital camera front as well as the handset camera front.

This is Mike.

I would be happy to address that.

We have a number of -- a dozen or more design wins on the digital still camera front, both actually with the one, two, and three megapixel chips.

We are rolling on the digital still camera front as well as we can reasonably expect.

Actually, better than we can reasonably expect.

On the mobile phone side, the sweet spot today, outside of Japan, on the camera front is still at VGA density and we have a number of design wins there.

We've got a couple of very high-profile design wins we are not talking about publicly, but we are shipping in the multiple millions of units now each quarter.

So we are very pleased with the -number of design wins that we have across those two platforms.

We are now out in the marketplace with our one megapixel sensor for mobile phones and we've secured a number of design wins now and we are moving into production very shortly on that chip as well.

So quite pleased with the progress.

Actually very pleased with the progress on the sensor front.

Great.

How about on the pricing front?

How have they been holding on?

Are you seeing pricing pressure or are they still --.

As the market moves on the digital -- it is a little different depending on the end market.

On the stand-alone digital still camera front as I will call it the pixel density race accelerates through 2 megapixel into 3 megapixel, we see quite a bit of price pressure on the departing densities.

So, in other words, today there is price pressure on the one and two megapixel chips for digital cameras, 3 megapixels is still selling at a nice premium and there isn't a whole lot of pressure.

On the mobile phone front it's a little bit different.

Actually what we are seeing is primarily due to the strong uptake in terms of camera penetration rate on mobile phones, I believe there is a supply constraint in the market currently and as a result pricing is stable to up for all the sensors that are going into mobile phones.

Great.

Just one last question on the CMOS sensor.

Do you expect to have the 4 megapixel coming out and do you expect the market will go to 4 megapixel this year or that could that be a 2005 product.

We are coming out with a 4-megapixel chip and we do believe the market will start to move to 4 megapixel towards the end of the calendar year.

Okay, thank you very much.

We would like to thank everyone for participating on the call today.

If you please bear with me, I need to repeat the Safe Harbor protection language.

During the course of this call we may have made forward-looking statements regarding the Company and the industry.

These particular forward-looking statements and all other statements that may have been made on this call that are not historical facts, are subject to a number of risks and uncertainties and actual results may differ materially.

For information on the important factors that may cause actual results to differ materially, please refer to our filings with the SEC including the Company's most recent 10-Q and 10-K.

Thank you for joining us.

Thank you, ladies and gentlemen.

This does conclude today's teleconference.

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