美光科技 (MU) 2004 Q4 法說會逐字稿

Good evening, everyone, and welcome to the Micron Technology fourth-quarter 2004 earnings teleconference.

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

I would now like to introduce your host for this call, Kip Bedard.

Sir, you may begin.

Thank you very much.

Welcome to Micron Technology's fourth-quarter and fiscal year-end 2004 financial release conference call.

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

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

This conference call, including audio and slides, is available on Micron's homepage on the Internet at Micron.com.

If you have not had an opportunity to review the fourth-quarter and fiscal year-end 2004 financial press release, it will be available on our website at 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 PM Mountain Daylight Time.

You can reach that by dialing 973-341-3080 and use a confirmation code of 5127647.

This replay will run through Thursday, October 7, 2004 at 5:30 PM, again Mountain Daylight Time.

A webcast replay will be available on the Company's website until September 29, 2005.

We encourage you to monitor our website at 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 this 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 you to 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 website.

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, Kip.

Let's repeat summary financial results for the fiscal year and fourth quarter, which ended September 2nd, for those listeners who may not yet have access to our press release.

For the fiscal year, net sales totaled $4.4 billion, and the Company recorded operating income of $250 million and net income of 157 million or 24 cents per diluted share.

That compares to net sales of 3.1 billion in fiscal year 2003 and a net loss of $2.11 per share.

The net sales increase of 42 percent year over year was facilitated by a 15 percent higher average selling price for semiconductor memory and an approximate 23 percent increase in megabits of memory produced.

Note that selling, general and administrative expenses, as a percentage of net sales, is reduced from 13 percent in 2002 to 8 percent in 2004.

For the fourth quarter, net sales totaled 1.2 billion, and the Company recorded net income of 94 million or 14 cents per diluted share.

Operating income came in at 125 million.

The income tax provision for the fourth quarter at 28 million is somewhat higher than recent quarters, and is a result of international taxation.

Our expectation is that income taxes will likely run about $20 million per quarter in fiscal 2005.

Gross margin for the year came in at 30 percent, an improvement from approximately breakeven in fiscal year 2003.

The fourth quarter's gross margin was 33 percent, holding up pretty well as compared to the third quarter, taking into consideration an approximate 5 percent decline in average selling prices quarter over quarter, and also taking into account a lower level of margin on products from our TECH joint venture.

Selling, general and administrative expenses were slightly lower in the fourth quarter, and will likely return to a run rate approaching 90 million per quarter in fiscal 2005.

The comparative reduction fourth quarter versus third quarter was a result of lower levels of taxes other than income taxes and lower legal costs.

R&D expense for the fourth quarter was 199 million, slightly higher than in recent periods.

The higher level was in part due to the expenses of wafer processing at the 300-mm line in Manassas, Virginia.

In the quarter, we shifted processing to our latest-generation 256-meg DRAM.

This hypershank (ph) 6F-squared DDR device, which uses utilizes copper interconnect technology, qualified for shipment to customers in September.

Future R&D expenses will vary significantly with the number of wafers dedicated to new device development and qualification.

From this next slide, you can see that cash flow provided by operations for the fiscal year was much improved over the prior year.

Future operating cash flows are significantly dependent on average selling prices, which for semiconductor memory are highly volatile.

Capital spending for fiscal year '04 came in below 1.4 billion, and fiscal year '05 is forecast to approximate 1.5 billion for capital spending.

As of year end, Micron had cash and investment balances exceeding 1.2 billion, and our debt-to-equity ratio remained below 20 percent.

I'll turn the commentary over to Mike.

Thanks very much, Bill.

Heading into fiscal Q4, we expected to see continued robust demand from all sectors, and for all intents and purposes, things occurred just about as expected with very few surprises.

The portfolio of DRAM products sold through to end customers with ease, in applications ranging from handheld consumer electronics products to high-end servers and everything in between.

Mobile phone applications consumed substantial amounts of our Flash memory, low-powered DRAM, CellularRAM and CMOS sensors, as we continued to advance our position in this key market.

Generally speaking, pricing for our components was stable throughout the quarter.

Our perception is that supply and demand are relatively balanced at this time.

Our finished goods inventory levels for all mainstream products are quite low, and should remain that way as demand strength in the vast majority of applications continues.

We don't anticipate material changes in the overall market environment in the current quarter that will significantly alter the current supply/demand balance.

PCs continue to be the largest consumer of our DRAM products.

We have a solid lineup of both DDR and DDR2 components and modules for the notebook, desktop and server markets.

Resurgent and seasonally strong PC unit sales and steadily increasing memory content per system are fueling sales of our DRAM products in this space.

We have put forth a tremendous amount of marketing and technology development effort to prepare for the DDR2 transition, and this effort has resulted in a solid lineup of 256-megabit, 512-megabit and 1-gigabit components on modules ranging in density from 128 MB to 4 GB -- without a doubt, the most complete product offering in the industry today.

For a variety of commercial reasons, sales of DDR2 memory have not ramped in the marketplace as aggressively as some industry watchers had projected.

We have been and will continue the practice of monitoring the market closely and dialing our production output to meet the specific product mix needs of our customers.

We have adequate capacity in place to accommodate virtually any DDR versus DDR2 demand scenario that might unfold over the course of the next several quarters.

Beyond PCs, in both the consumer electronics and communications space, we are seeing sufficient demand strength in the market to absorb our expanding production output.

The most prominent of these sectors, the mobile phone market, is experiencing both solid unit growth and expanded functionality this year.

With industry-wide infrastructure enhancements taking hold, Mobile handsets are moving upmarket such that the total available opportunity for Micron's products is becoming increasingly interesting.

This is specifically resulting in increased memory content per mobile platform and growing camera phone penetration.

On the memory front, we have positioned our CellularRAM, low-powered DRAM and NOR Flash products in the mobile handset space.

This business is continuing to grow, as the market expands and as Micron's share of the available market increases.

We are adding an additional memory product family to the portfolio, NAND flash, that will also find a place in mobile phone platforms.

In fiscal Q4, we started sampling our initial 2-gigabit NAND component, and feedback from customers on device performance has been positive.

Our plans to commence commercial shipments of this device by the end of calendar 2004 remain unchanged.

We now have both VGA and 1-megapixel CMOS sensors in volume production to support the mobile phone camera business.

These chips are being well-received by the mobile phone industry for their image quality, efficient power consumption characteristics and low cost relative to CCD.

We are continuing to enhance the product roadmap by rolling out cost and format reductions of these chips, as well as higher-density mobile phone camera chips in coming periods.

We are pleased with the position we have established with the mobile phone makers, and this is attributed to a solid memory and CMOS sensor product portfolio and strong customer support infrastructure.

We're moving through the first quarter of fiscal 2005, and from a demand standpoint, the market for our products continues to be quite favorable.

Micron's product portfolio and position in the marketplace is better than ever.

We are maintaining a solid standing in the high-volume computing platforms, and leverage the core technology to enhance the position in consumer, communications and other emerging applications.

The environment, both external and internal, is quite exciting, and we thank you for your continued interest and support.

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

Thank you, Mike.

We would now like to take questions from callers.

Just a reminder -- if you are using a speakerphone, please pick up the handset when asking a question so that we can hear you clearly.

With that, we would like to open up the phone lines.

(OPERATOR INSTRUCTIONS).

Michael Masdea, CSFB.

Can one of you just run us through real quickly what actual bit production increased and then bit shipment increased in the quarter?

Sure.

We were up low double digits on bit production, and the growth rate in sales was slightly above that.

And then can you -- maybe Mike can run us through kind of what the customer dynamics -- because it sounds like there weren't any surprises.

But maybe, given what we have seen in some of the other end markets recently, maybe you can give us an update on what your customers are thinking about next quarter.

Sure.

In next quarter, which will be calendar Q4, fiscal Q1 for us, we're looking at from the PC space probably megabit consumption growth rates of around 15 percent, if I can kind of take an educated guess or educated rounding of what customers are telling us.

And that's roughly in line with what would seasonally be occurring.

And my perception is that was probably about the same growth rate that we saw in calendar Q3, as well.

In general, in the PC space -- which continues to be the largest consumer of our DRAM megabits, anyway -- things are happening about like we would have expected, from a seasonal standpoint.

In other markets, whether consumer electronics markets, also mobile phones, also -- things are occurring about as expected.

We're continuing to get the content boost that we had been expecting in all of these products, and we are continuing to roll out products to take advantage of that, as well.

So, generally speaking, Michael, from an external market standpoint, everything is pretty much happening as we thought it would, and as we commented -- like we thought it would at the last conference call.

And I guess, relative to that, how much can you guys grow your bits next order?

We are starting our guidance at low double digits again.

And then just the last question is on the inventory front, and what you're seeing there, what it looks like in the channel and in total.

On finished goods inventory, we are pretty tight.

We're at -- ended the quarter at less than two weeks of supply.

We're currently in that range right now.

We've added to it a little bit in the first couple of weeks of the quarter, just to get inventories back in line.

My perception on the customer front is very little inventory on hand.

We're still getting hit pretty hard by most of our customers on a daily basis for deliveries, and that would be a pretty strong indicator that there's not much inventory out there.

I couldn't really speak for the other suppliers, but again, to summarize, our inventory is less than two weeks of supply.

My observation is that our customers are pretty much out of inventory, as well.

But prices looking pretty stable?

Prices are -- DRAM prices, which is what most people focus on, are stable to trending up.

We did, midmonth, realize a small price increase from the majors, and things look like they are probably on cue for another price increase here in a couple days.

Glen Yeung, Smith Barney.

Maybe taking a bit of a longer-term look at your CapEx for fiscal '05, maybe just run through the rationale for taking it up, given what seems to be a growing consensus that we may be entering into oversupply.

So maybe running through how you intend to spend it, what your wafer outs looks like next year and what your bits out look like for fiscal '05, as well?

There's a lot of questions in there.

I'll tackle the capital, and then perhaps Kip can talk about some of the wafer and bit.

On capital, I think, as most people are aware, we are ramping our 300-mm facility over the next six to eight months.

And actually, our capital is a little bit front-end loaded.

So your comment about whatever happens in the latter half of the year, of course, is difficult if not impossible for us to predict.

But we have made the commitment to bring that 300-mm facility up, and in fact a lot of equipment is being or has been delivered, and so for our fiscal year we are a little heavier in the front end of the CapEx.

But again, that is a pretty big initial expenditure, and that's why you're seeing our CapEx move up a little bit.

And then from an industry bit growth standpoint, if you look at the various analysts who study that -- IDC, iSupply, et cetera -- they are anywhere in the range of low 40's to low 50 percent production bit growth '05 over '04.

We will be in that range, and of course some of that is subject -- exactly where we are in that range is somewhat subject to how we allocate wafers to non-DRAM products.

Ted Parmigiani, Lehman Brothers.

Just trying to understand some of your costs, as well as the comments on ASPs.

You talk about memory ASPs declining 5 percent.

How should we understand that, given stable to higher contract prices quarter on quarter for you guys?

Is there something going on in mix there?

And then also, if you could talk about two more things -- one is the overall mix along the lines of memory versus non-memory, and also sequential reductions in cost, what you saw percentage-wise in the prior quarter, as well as what you expect to achieve in this current quarter.

I'll address the ASP question and the product mix question, as well.

I think, on the 5 percent reduction versus stable to increasing contract prices, we're mixing periods here.

The 5 percent ASP reduction occurred in overall in our entire portfolio fiscal Q4 versus fiscal Q3.

The stable to increasing prices that I am referencing are today's situation.

So we're in fiscal Q1; actually, we are 3.5 weeks through fiscal Q1, and prices did pick up slightly the first couple weeks of the quarter, and they are probably due to tick up again here over the course of the next couple days.

So, two different periods on the pricing question that you had.

With respect to the product mix, generally speaking, from a product mix standpoint about a third -- somewhere between 30 and 35 percent of our overall wafer starts -- are in the non-core DRAM area, and that is consistent with the messaging that we have been stating.

And the biggest case of the non-core DRAM, of course, would be the synchronous DRAM followed by the CellularRAM and then the Flash memory and the CMOS sensors, as well.

Did we get all your questions answered?

No.

Actually, on the production costs, you had talked last quarter about expecting to achieve a percentage reduction in production costs.

I was wondering if you had met that goal, and also what you could comment on going forward here in the current quarter.

To part of that question, we indicated in opening comments that the margin that we realized on products out of our TECH joint venture was less in the fourth quarter than that which we had realized in the third quarter.

Had that held level, gross margin would have stayed constant Q4 relative to Q3, so you can get a relationship of the cost profile on our core memory, just by looking at that gross margin staying constant and seeing the characterization of the fourth quarter versus third quarter ASP decline.

One of the things I would note there is that the -- remember, on the cost profile for our advanced product, as Mike just mentioned, that only constitutes a certain percentage.

So we have certain products of our lower density or legacy, or however you want to categorize it, where we really don't make that much cost improvement.

Of course, the margin is strong on those, as well.

But you've got to keep that in mind, that the overall cost profile reductions that would occur -- those primarily relate to the advanced -- the mainstream core DRAM, and they don't really relate to some of these other product.

John Lau, Banc of America.

I was wondering if you could go over some of the dynamics in the CMOS sensor area, especially for the digital still cameras, and hence what is the particular density (ph) trends over there?

And then, you had mentioned that inventory was low.

How about inventory for the CMOS sensor product lines?

And then, finally, is CMOS sensors still about 15 percent of the wafer starts?

I'll address them, and Steve might pipe in here, as well, with his input.

On the DSC front, my observation is the sweet spot density today is probably 3 megapixels, shifting to 4 megapixel.

One of the things that has occurred in the DSC market is that CCD technology has held onto a larger percentage of the business than Micron and other CMOS makers had anticipated probably a year ago.

And as a result, the opportunity that has been realized for CMOS sensors in the DSC space has not been nearly as significant as we would have expected.

The bottom line is that the vast majority of our revenues today are coming from the mobile phone space, and the vast majority of our resources today are working on chips for both mobile phones, automotive applications and other emerging applications, as opposed to DSCs.

Now, that doesn't mean we're ignoring the DSC market, but what we are doing is we're embarking on a strategy to leverage the same designs, basically, for both the mobile phone market and the DSC market.

You had a question also about --

Inventory.

-- inventory.

Its low; it's very low.

On the mobile phone area, as you know, we have a VGA chip in high-volume production today.

Our inventory level on that -- I don't have the specific number at the tip of my fingertips, but it's quite low.

It's in a week or two weeks of inventory, very low.

We have introduced a reduced format of the VGA, a 1.6-inch VGA chip, and we are just in the initial ramp-up stage right now.

We have virtually no inventory of that.

And as I mentioned, we have also got a 1-megapixel mobile phone sensor in the marketplace, and we have got virtually no inventory of that.

So any inventory that we have accumulated has been 1, 2, 3-megapixel DSC chips, and I already talked about the current state of the DSC market.

And then finally, you had mentioned in the past that CMOS sensors was about 15 percent of your wafer starts.

Is that still about right?

It's trending that direction.

We're still in the single-digit range.

Adam Parker, Sanford Bernstein.

I had a couple questions here.

Sort of long-term, if international growth in PCs continues to exceed the US growth, how do you sort of plan on increasing your Asian exposure, or is that part of your long-term plan?

It's absolutely part of our long-term plan.

We're currently generating about a third of our revenues in the Asia-Pacific region, and as a subset of that, what we consider to be greater China -- which is Taiwan, Hong Kong and China itself -- about one-fifth of our revenues or about 20 percent of our revenues are coming from that region.

What we're doing is replacing more resources there to support the growing customer base -- both the multinationals that are leveraging the low labor cost areas in China, in particular, as well as the indigenous Chinese manufacturers.

Without question, we're the largest memory supplier to the largest PC manufacturer in China, Lenovo, and the second-largest in the region, we also have a very strong presence with, as well.

So what we're doing is that if we use the Lenovo relationship as a model, we are attempting to replicate that with both PC makers, mobile phone makers and other electronics equipment makers in the region.

So we are very well situated in the Asia-Pacific region today, generating a substantial amount of our revenues today and growing resources such that we can continue on that same kind of trajectory.

Just a couple of quick housekeeping items.

What percent of your revenues -- I think you may have said this, but I didn't get it -- were from non-DRAM during the quarter, and what was legacy DRAM during the quarter?

If you consider legacy being basically SDRAM, figure about 15 percent of the wafer allocations, in that range, another 15 to 20 percent would be non-DRAM-related or specialty DRAM, which would be inclusive of image sensors, Flash, pseudo SDRAM, RLDRAM, mobile RAM, et cetera.

Kip, that's wafers, right?

That's wafers, yes.

We tend not to break out the revenue numbers from those.

One other thing.

Did you say anything about -- I didn't hear you say anything about the R&D guidance for '05.

I heard you say SG&A.

Do you have a range for us for R&D for the year?

The 199 was what we reported for the fourth quarter.

Q1 -- actually, all the quarters through '05, you can use the 175 reference point as pretty solid.

And then, the tax rate for the year -- obviously, it is contingent on making money and so forth, but is there any reason why we should expect it to go above what you just reported?

Or you cannot guide on the tax rate at all, because it's too contingent on earnings?

If we thought of a tax dollar amount is probably the better way to think of it, as it's a function of our international operations, most specifically Japan and Italy.

If you use the 75 million, 80 million number for the year, that's a good way to think of it, because the net operating loss will largely offset the US profitability.

And the last thing -- I apologize;

I got the press release after the call started, but I did not see how the share count declined sequentially.

Can you just remind me how that happened?

The number of shares used in the earnings-per-share calculation fourth quarter of '04 was 645 million for basic, 701 for the diluted calculation.

And for the year, it was 641 and 645, remembering that for the year, the convert is anti-dilutive, so it's not included in that diluted calculation for the year.

Maybe I misread it.

I thought Q3 versus Q4, the diluted share count came down.

But maybe I'm not up to speed there.

Well, we'll have to work that one off-line with you.

It's got to be a function of the interest add-back associated with the convertible debenture.

One last thing.

What was the total contribution that you guys did for the full fiscal year on the employee profit-sharing, just in millions of dollars?

We don't have that at our fingertip, Adam.

Well, remember that it runs 10 percent of our net income.

So in aggregate, if you wanted to look at it that way -- of course, for the first two quarters there wasn't anything, and the for the second two quarters we had profitability.

But roughly, it's in that neighborhood.

The 10 percent program would aggregate 185 million.

It's 10 percent of the 185 million in the third and fourth quarter.

Andrew Root, Goldman Sachs.

The impact of 6F-squared qualification on your costs next quarter -- does that result in a bit of a boost to your gross margin by putting those products in inventory now?

Remember that we are continuing to convert more products to 6F-squared, so keep in mind that we first attack the mainstream runner and DDR, and then we move through it as we convert those.

So kind of a general way to look at it is we have a lot of headroom left on the conversion of the 6F-squared -- the devices to 6F-squared.

I don't have it right at the tip of my fingers, the impact specifically (ph) for the next quarter.

But obviously, for the year it's going to continue to make a difference every quarter that goes by.

And then, if I read the press release correctly, your megabits shipped of DRAM, excluding Flash and image sensors, your growth last fiscal year was 23 percent in megabits.

Is that right?

That sounds a little bit lower than I would guess.

Do you have an estimate of what fiscal '05 will look like?

Yes.

Just before, we said the range -- we will be pretty much in range with the industry, at somewhere between low 40's and low 50 percent for DRAM.

And I might also mentioned that bit growth number also includes Flash but excludes image sensor.

Because the 23 percent -- that range is sort of a similar range, I guess, you were using through most of last year and ended up quite a bit lower.

Is it just the pull of the incremental profitability of the products will just keep moving it in that direction?

Correct.

As Steve mentioned earlier, and Mike reiterated as well, some of the products -- if we continue to allocate more wafers, for example, to 64 and 128-meg SRAM, the ability to shrink those and drive bit growth out of that is more difficult than if you are allocating more wafers to, say, 256, 512 or 1 gig.

So some of that is reflective of the actual wafer allocation.

I guess maybe I'm asking the question in a tortured manner.

But if things basically stay the same, is there any reason to believe that DRAM bit growth doesn't end up at 25 percent in the next 12 months?

Again, right now we're looking at pretty close to industry average, but if we decided to make a pretty major wafer allocation change, that numbers could be higher or lower.

The thing to remember is that, as I just mentioned, we have headroom because we're continuing to convert more products to 6F-squared.

We have headroom left on our 20 series yields (ph), if you will, as we continue to optimize those.

And then let's not forget we have a 300-mm fab that's coming on that are wafers that are not even in the production bit count today, that will ramp during that period.

So we would expect to get bit growth from all three of those sources, which is a little bit different than what happened in fiscal '04.

And then, the final question is there is no provision in this release for any Department of Justice settlement or charge?

One of your competitors obviously took a big charge.

I knew you guys are cooperating, but if you could give an update there, that would be great.

Well, as we said, we don't anticipate any charges or penalties against the company from the DOJ.

David Wong, A.G. Edwards.

You said that the PC demand package was pretty much as expected through the quarter, but can you give us any sort of finer detail, in terms of desktop status, notebooks and perhaps consumer versus commercial demand?

Only qualitatively.

The notebook business is quite strong; the server business is quite strong.

But I really couldn't put any numbers behind it.

It just seems that the screams are louder for our notebook memory and for our server memory than they are for desktop.

And any insight between consumer and commercial?

We don't have much visibility there, unless I specifically go out and seek it, because it's by and large the same products that we -- the same components and modules we manufacture for either platform are flexible, fungible between the commercial and the consumer desktops.

Tai Nguyen, Susquehanna Financial.

Could you comment on the communication market?

I remember last quarter you were focusing a lot of your NOR Flash towards that market, and it seems like you're kind of shifting now the NOR Flash towards the handset market.

Yes.

We're not shifting the NOR flash to the handset market.

We've got exposure in both mobile handsets communications as well as even some consumer electronics gear with our NOR flash product offering.

My observation on the communication -- I presume what you are talking about are the communications infrastructure gear like routers and switches and so forth, base stations.

And my observation is that business is kind of plugging along at a reasonably healthy growth rate, nothing extraordinary occurring there.

Then as far as for next quarter, are you expecting to continue with the high rate of the R&D, or are you expecting the wafer for 300-mm to come down?

Well, Bill already answered that in an earlier comment, and we finished last quarter just under 200 million.

And I think he said that you use basically 175 million as a good reference point going forward.

So it will come down, as those products convert from R&D into production and then drop through the cogs.

Krishna Shankar, JMP Securities.

To what extent were ASPs down in Q4 because of the focus on the legacy SDRAM business, and could that reverse in Q4 as you focus more on DDR and DDR2 PC-centric DRAM?

I presume what you are referring to is some market price pressure on the low-density SDRAM products.

And to be honest with you, in our fiscal Q4, we did see some price pressure in the low-density SDRAM area, but that, in aggregate, had very little impact on our overall average selling prices for DRAM.

It was primarily fairly steady -- I would almost characterize it as stable to slightly declining pricing in the PC space that resulted in the 5 percent quarter-over-quarter ASP drop.

In terms of fiscal Q1, as I mentioned, we are 3.5 weeks through it.

On an aggregate basis, I think our shipments out the door today, the ASP is slightly higher than what it was in fiscal Q4.

But again, the big swing factor through the whole quarter is going to be what DRAM prices do, particularly in the PC environment.

And the trend, as I mentioned, is currently up.

And in terms of DDR2, what percent of your mix do you think will be DDR2 in Q1?

And for the industry as a whole, what do you think that will be going into next year?

I think that going out of this calendar year, we will probably be at about 15 percent of our total shipments being DDR2 vintage.

And we're trying to peg that to be roughly in line with what we think the market is going to be.

And the final question -- can you give us a sense for how quickly the 300-mm fab ramps up, and the impact of that on your gross margins?

Well, with respect to gross margins, I don't have that data right at my fingertips, but the ramp is pretty steady from this point forward up through the summer of '05.

So there is no dramatic changes that would occur.

Obviously, it's a negative on the gross margin at the moment, but we have said we think it will become basically a positive on the gross margin after a couple of quarters.

Matt Gable, Calypso Capital.

I might have missed this.

What did your fully-loaded costs do sequentially in the August quarter, and is there any guidance for the current November quarter on a sequential basis?

Actually, we did cover that before.

And through Bill's comments, you could figure that it was probably about a 5 percent for Micron-controlled production, for our costs for the quarter.

And going forward, as Steve mentioned, we haven't given a specific cost target number.

But the three areas that we do have working in our favor, of course, outside of what pricing might do, is we have headroom on 6F-squared.

We have headroom on 110-nm maturity factor, and we still have 300-mm in front of us, as well.

But also remember that the number bill referenced is an aggregate number.

We have lots of products that are in that revenue flow that we really don't have much cost reduction on at all, because it's maybe a higher margin part, but it's a lower density or legacy device that's running on a process that's not utilizing the most advanced -- that we use to reduce the cost per bit.

Nimal Vallipuram, DRKW.

I have a few questions here.

First of all, congratulations on a good quarter here.

The first question is to Mike.

If you look at the last quarter, third calendar quarter, it was pretty weird in terms of DRAM pricing, because the expectations of the PC market just at the beginning of the quarter was pretty high.

And at the end of the quarter, it turned out that at least July and August were pretty weak.

But if you look at the DRAM prices, at least the contract prices, did not move much just throughout the third calendar quarter.

Can you give us an idea why that is the case?

Did you feel that's the same way, too, and why it is the case?

And also, in your comments you don't seem too enthusiastic that in the fourth quarter, given that there is always some seasonality, that you don't think that the prices are going to seriously move up.

Am I reading that right or not?

I'm not in the business of predicting what pricing is going to do.

To be honest with you, I think what occurred in our fiscal Q4 and what has occurred thus far in the first couple weeks of fiscal Q1 are absolutely perfect, with respect to pricing.

Our customers can take comfort in relatively predictable -- in this business, anyway -- relatively predictable DRAM pricing, relatively stable DRAM pricing.

They can plan a pretty rich portfolio from a memory content standpoint, and I think it's absolutely ideal.

These days, by the way, or at least in the last couple of years, and what I expect to be the pattern going forward, is that seasonal strength in calendar 4Q is really -- it starts in probably the July-August timeframe, and it really goes through November.

So it's kind of a second half of calendar Q3, first half of calendar Q4 seasonal strength.

We are right in the middle of it right now, and that's absolutely what I am observing.

In terms of the price being relatively flat to slightly going up, that's just a byproduct of a relatively balanced supply/demand environment, again, which in this business is ideal, from our standpoint.

So we couldn't be happier with the way the market is performing right now.

Just a follow-up on that.

Does it mean that any hidden variable from the supply side as some like (ph) some of your companies are having yield problem (ph) or anything of that sort continuing from what happened in the beginning of this year?

I think that, in aggregate, these things all tend to balance each other out.

So clearly, some of our competition has had yield difficulties.

But in the same breath, some of our competition has brought on more capacity.

And as you go through these periods of time, I don't think it's any one factor, and it tends to balance it all out in aggregate.

Just a follow-up question.

You said something about -- you gave a guidance for the SG&A.

And I might have missed it in the initial comments, but the SG&A this quarter was pretty low compared to last quarter.

Is that a number which you plan to have going forward?

No, the number which we indicated is it will probably move back to about a $90 million per quarter level.

Okay, sorry.

That's what I missed, I suppose.

Any reason why it was 75 this quarter, why it was so low?

We had indicated that we had slightly lower legal costs, had slightly lower level of taxes other than income were the primary factors causing a reduction this particular quarter.

Oren Hershman, OH Investment Group.

Some of this you covered, to some degree.

But perhaps, if you would be kind enough to do a recap, in terms of the joint venture and the effect on gross margins this quarter, and how it should affect gross margins this quarter and the coming quarter, the quarter ending now and the following quarter, can you just review that again quickly?

What we indicated is that the pricing arrangement with the TECH joint venture is a lag quarter pricing.

It had the effect of the margin which we realized on the product out of TECH in Q4 being somewhat less than the margin we had realized in Q3.

As part of the forward-looking, you'll have to take, again, you have to draw your own views as to what ASPs are going to do before you can make your own assessment there.

Did you actually quantify the impact that it had on margin?

The only reference that I'd given is that, had the margin on the TECH product stayed at the same level as Q3, overall gross margin would have been 35 percent, consistent with Q3.

And in terms of actual pricing in the CMOS sensor market, can you talk a little bit more about that, and if the shift to the higher pixel counts are beginning to take pressure off the pricing competition?

The specific answer to your question is that the higher pixel count chips -- for example, the 1-megapixel versus the VGA chip -- the selling price is higher.

There is no question about it.

I'm a little bit hesitant to say it takes pressure off the price, because we will continue to see competitors at every density level on every chip that we manufacture.

Specifically, the VGA pricing has been relatively stable, although we plan to get quite aggressive with our VGA chip pricing, because we have got what we believe is the lowest cost structure and the best image quality in the industry.

So we are going to get quite aggressive with the VGA pricing strategy going forward here.

And we will continue to price all the products in the portfolio competitively in the marketplace.

Did you indicate whether that business grew revenue-wise this quarter versus last quarter?

I couldn't quite understand that question.

Did you indicate, just on the CMOS sensor side, whether you were up sequentially in terms of revenue?

No, we have not.

Ben Lynch, Deutsche Bank.

A lot of your competitors have had very high operating margins for several quarters now, some of them in the 30 to 40 percent range.

And as you know, DRAM rarely stays very profitable for very long.

How worried are you that Micron's window of opportunity for making decent margins and actually boosting your cash position is maybe passing for this upcycle?

I think you have to keep in context what we have been trying to achieve.

It is true, we could have sacrificed some of the development in the other product areas in order to get better margin in the -- in particular, by the way, in the primary largest-volume DRAM part that goes into the market.

So if you look at one or two of our competitors -- by the way, it's hard to get -- unless they kind of report in a pure form, it's tough to know exactly what that margin is.

But even having said that, in other words, if we wanted to put all of our wafer starts on a 256-meg DDR and not make anything else, then obviously we could take advantage of an improved DRAM market, and are margins would be quite a bit higher.

But the difficulty is we would then not be very well positioned, as you already just stated, when the price falls out on the 256-meg DDR, whenever that may be.

We don't know when it's going to be, but whenever it occurs, obviously we would be much more impacted by it.

So we have specifically been trying to not only position the product portfolio on some of the devices that we are calling legacy, but they still may be fairly advanced devices, like the synchronous DRAM, as we continue to shrink those and move them forward, or some of the specialty devices like pseudostatic RAM, which is going into the non-computing environment, which has been a very, very profitable part.

So, regardless, as we diversify to some degree, it does require more resources to do that, and we do incur some higher costs, at least in the earlier stages.

So, if I kind of sum this all up, I've said this before, if you look at the DRAM alone and you don't count the resources that we are applying toward some of these other products -- which obviously today may not be quite as good a margin as the DRAM, but ultimately will probably have a better margin than the DRAM when the margin in the DRAM does decline -- even having said all that, we still have quite a bit of headroom in our 110-nm yields.

We have headroom as we continue to convert more to 6F-squared of the main DRAM runners, and you can't ignore that we at least historically have been the process of ramping the 300-mm facility.

So we think that that has bottomed, if you will, in terms of the drag on the Company, and now we have an improving scenario going forward on 300-mm.

So we essentially paid that price, historically, and it actually will be an improving scenario for us on the contribution to the Company.

So we feel pretty good about where we are at.

Just a follow-up, please.

Steve, you might not want to state specifically, but how much headroom is there in 110-nm yields?

And also, of mainstream DRAM today, how much of wafer adds are on 6F-squared technology?

Well, let me go back to that.

On the 20 series yields, it depends on the device, of course.

And when you say yield, by the way, it depends on whether your talking about manufacturing yield or wafer yield, cumulative (ph), et cetera.

I would say in aggregate, on the yield front, we can probably get somewhere in the neighborhood of 10 to 15 percent more on the cumulative side, which is obviously highly leveraged for us.

With respect to the 6F-squared, we only have about 40 to 50 percent of the product running on 6F-squared.

And obviously, as we move through what we call our 90-nm and our 78-nm device, we would expect that to essentially go to 100 percent.

Tonya Saul (ph), Ragen MacKenzie.

I just had a question here.

We've been hearing a lot of indications from various companies on weakening consumer demand or weaker than normal calendar Q4.

And it sounds like you are not seeing that, that you are basically seeing there is demand out there for computers and the other consumer items like digital still cameras that you guys sell into.

And can you just kind of talk about that for a second?

I am quite confident in the end products that we have exposure to, which would be, in order of significance, PCs, cellular phones, communications infrastructure equipment and consumer electronics gear.

I'm quite confident that in the customer base we're dealing with, there is sufficient demand to absorb everything we're going to produce -- very simply stated.

Clark Fuhs, Fulcrum Global.

The wafer outs -- where are they currently?

We were up a little bit this quarter over last quarter.

We are running high 40,000 outs a week -- 300-mm equivalents.

High 40,000?

That has not reached the 50K mark yet?

I has, at times.

Again, it depends on product mix, and it has been moving around that number for a while.

And can you remind us what your summer '05 300-mm level is going to be?

Yes, we think on -- well, for a 300-mm equivalent, of course, it's going to be somewhere between 10,000 and 15,000.

And you have a goal of your internal cost per bit --

Clark, just one qualification on the 300-mm.

I didn't give you the timeframe.

That's going to be per week.

Right.

I just wanted to be clear.

You have an internal cost per bit reduction plan of about 10 percent per quarter; you stated that before.

Did the fourth quarter meet that goal, or was it more like high single digit?

The timing that we have given earlier in the call was when you take out the effect of the TECH joint venture, it essentially tracked with ASP reduction of 5 percent.

That's in aggregate; that's not the devices that we actually would be focused on in solely reducing that cost.

So really, the way to look at it is what was the bit efficiency gain for the 256-meg 512 DDR, in other words?

I'm kind of looking at more what the underlying --

It's probably -- I don't have the number.

I think it's probably in line with what we have done historically, which is around that 10 percent.

Right, I would guess actually more like high single-digit, but okay.

Just a housekeeping -- depreciation expectations for Q1?

We ran 309 in Q4, and you're going to be running 325 to 330 per quarter in '05.

And what percentage of wafer starts now are 512-megabit?

We can add those up for you here real quick.

We are running about almost close to 20 percent on starts today.

And as far as 95-nm technology, you mentioned a few weeks ago that you had started that production, and it was kind of running a little ahead of plan.

But what percent wafer starts would you expect by calendar year end, and say by the middle of next year?

We're on target with what we have said before, and about the best guidance we have given you is just so that we don't tip off our competitors too much, is we will have a couple percent aimed at the 90-nm transition as we exit the calendar year.

Paul Lemming (ph), Soleil Securities.

I was wondering if you have looked at what profitability would have been if you had not made the move over the last couple of quarters to shift wafers over to the nontraditional DRAM.

Just as you do a snapshot on the quarter and look at wafer capability, did the transition over to image sensors, over to Flash, help or hurt margin in aggregate in the August quarter?

No, we never did.

However, let me just make one comment along those lines.

The ASP probably would have been lower, because we would have pumped out a lot more mainstream DRAM.

Fair enough.

And I think (multiple speakers).

I'm not sure we are capable of even doing that calculation.

And I understand there's a lot of pluses and minuses in doing it.

And actually, all my other questions have been answered.

I think we have time for one more question.

Ted Parmigiani, Lehman Brothers.

Just a follow-up on the CMOS image sensor business.

You talked about, you believe, having the lowest cost structure out there.

And if I look at your largest kind of fabless pure play competitor and the margin structure, the gross margins they are currently getting for their product line, can you talk about what maybe your short-term and long-term hurdle rate or kind of goals are, with respect to gross margin in this business, and what we should be thinking about, in terms of modeling a kind of margin for this business going forward?

Are you talking about Mandy (ph) or Hynix?

Excuse me?

Are you talking about CMOS imagers?

Yes, exactly.

Oh, okay, because Hynix made some CMOS imagers, of course.

I'm talking about the fabless player out there.

Oh, fabless?

I'm sorry.

You're talking about Omnivision.

Let me get the question straight again.

You want to know what we think is a good (multiple speakers) --

You have a low-cost structure, and presumably you are going after some market share with not only the technical merits of your products but also with price, to some extent, in the short term.

But what should we be thinking about longer term, or what are you guys thinking about longer term, in terms of what this business can mean, in terms of gross margin?

Well, we think that 50 percent -- I don't know if -- it depends on the exact product line and where we are seeing the competition, but clearly somewhere between 45 and 55 percent gross margin, we think, given our cost structure, we should be able to achieve.

Thank you very much.

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

If you will 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, everyone.

This does conclude today's teleconference.

You may disconnect all lines at this time, and have a wonderful night.