Kopin Corp (KOPN) 2003 Q4 法說會逐字稿

Good day and welcome, everyone, to the Kopin Corporation fourth-quarter and year-end 2003 financial results conference call.

Today's call is being recorded.

Please let me remind everyone that the replay of this conference call will be available from 8:00 PM Eastern time through Monday, February 23, by dialing 719-457-0820 and entering confirmation code 757591.

You may also access an archived version of the call on Kopin's website at www.Kopin.com.

With us from the company is the President and Chief Executive Officer, Dr. John C. C. Fan, and the Chief Financial Officer, Mr. Richard A. Sneider.

At this time, I'd like to turn the call over to Mr. Sneider.

Please go ahead, sir.

Thank you.

Good afternoon, everyone, and thank you for joining us for Kopin's fourth-quarter and full-year 2003 conference call.

I will begin by reviewing our financial results for both the quarter and full year.

John will then update you on our recent operational achievements and share our outlook for the coming year.

Then we'll take your questions.

Before we begin, I want to remind everyone that during today's call taking place on Tuesday, February 17, 2004, we will make forward-looking statements as defined in the Private Securities Litigation Reform Act of 1995.

The statements are based on the company's current expectations, projections, beliefs, and estimates, and are subject to a number of risks and uncertainties.

Potential risks include but are not limited to demand for our CyberDisplay and III-V products, market conditions, the company's ability to ramp up production at its manufacturing facilities, and other factors discussed in the company's 10-K for the year ended December 31, 2002, and 10-Q for the three months ended September 27, 2003, and other documents on file with the Securities and Exchange Commission.

The company does not undertake any duty to update any statements made during today's call.

If you've not received a copy of today's news release, it is available on our web site, www.kopin.com, or you can contact Sharon Merrill Associates at 617-542-5300 and a copy will be sent to you.

Turning to our financial results, total revenue for the fourth quarter of 2003 was 21.1 million, up 21 percent from Q3 revenue and in line with our expectations of a 20 percent sequential increase.

Year-over-year revenue for Q4 of 2003 increased approximately 28 percent from the comparable period in 2002.

III-V revenue in the fourth quarter of 2003 was 8.8 million, up 46 percent from 6 million in the third quarter of 2003 and up 56 percent from the 5.6 million in the fourth quarter of 2002.

We estimate that HBT sales for wireless LAN applications were about 10 percent of HBT sales, a number (ph) that has remained consistent throughout 2003.

Our CyberDisplay products generated 12.4 million in revenue in the fourth quarter of 2003 compared with 11.5 million in Q3 of this year and 10.9 million in Q4 of 2002.

This is the highest display revenue achieved in a quarter.

Net loss for the fourth quarter of 2003 was 614,000, or 1 cent per share.

This compares with a net loss of 3 million, or 4 cents per share, in the sequential third quarter, and a net loss of 15.1 million, or 22 cents per share, in the fourth quarter of 2002.

Net loss for the fourth quarter of 2002 included an unrealized loss of 10.2 million from the write-down of Kopin's investment in Micrel Semiconductor.

Cost of goods sold in the fourth quarter of 2003 represented 75.9 percent of product revenues versus 85.8 percent in Q3 and 75.2 percent one year ago.

The improvement in the gross margin over Q3 of 2003 reflects the absorption of fixed cost from the increase in HBT unit sales.

Research and development expenses were 4.4 million, or 21 percent of revenue in fourth quarter, compared to 3.3 million, or 19 percent of revenue in the prior quarter.

We project R&D expenses to be in the range of 15 to 20 percent of revenues through Q1 of 2004.

Selling, general, and administrative expenses in Q4 2003 were 2.2 million, or 11 percent of revenue.

Fourth-quarter SG&A expenses were favorably impacted by about 300,000, which resulted from the truing up of budgeted expenses to actual.

We would expect Q1 2004 SG&A to be the range of 2.7 to 3 million, which reflects certain normal seasonal expenses.

For the full year ended December 31, 2003, total revenue was 76.6 million versus 76.8 million for the same period in 2002.

Revenues from III-V products was 32.9 million compared with 32.7 million for the full year 2002.

Revenue from CyberDisplay products was 43.6 million, compared with 44.1 million in the same period a year ago.

The year-over-year decline in CyberDisplay revenues is related to the ASP declines in the monochrome CyberDisplay business, which were partially offset by greater volume shipments to existing and new commercial and military customers.

Kopin's net loss for 2003 was 6.9 million, or 10 cents per share, based on 69.5 million shares outstanding.

The net loss for 2002 was 31.9 million, or 46 cents per share, based on 69.3 million shares outstanding.

Included in the 2002 net loss was the cumulative effect of a change in accounting of 12.6 million, and a 10.2 million write-down of the company's investment in Micrel.

Our financial position remains very strong.

As of December 31st, Kopin had cash and marketable securities of 120.3 million compared with 118 million as of December 31, 2002.

In addition, Kopin owned approximately 400,000 shares of common stock of Micrel, which at December 31, 2003 was worth approximately 6.5 million.

Inventory at December 31, 2003 was 5.9 million versus 4.8 at the end of 2002.

The increase reflects our sole source supply -- HBT supply agreement requirement to maintain certain inventory levels.

Our 10 percent customers for 2003 were Samsung at 32 percent, Skyworks Solutions at 21 percent, and JVC at 12 percent.

For the fourth-quarter, band-gap (ph) sales were approximately 55 percent of HBT sales, and for 2003, band-gap was approximately 62 percent of HBT sales.

Cash flow provided by operating activities in 2003 was approximately 6.5 million.

DSOs were running about 30 days for the quarter compared to 50 days in Q3.

We would, however, expect 50 days to be a more normal DSO period.

For the current quarter, companywide fab utilization was approximately 70 percent.

CapEx for the year was 5.4 million.

Depreciation and amortization was approximately 9.7 million.

We would expect CapEx in 2004 to be in the range of 5 to 7 million.

As a result of normal post-holiday seasonality, both CyberDisplay and III-V revenues usually decline sequentially in Q1.

However, our initial shipments of some of our new CyberDisplay products to new customers are expected to counteract the seasonal drop.

Based on this expectation, as well as current business trends, market conditions, and the outlook for our end-market, we anticipate the total revenue for the first quarter of 2004 will be flat to slightly up, which will be an increase of about 15 to 20 percent year over year.

I'll now turn the call over to John.

Thank you, Rich.

Before I begin our review of the quarter and the year, I just wanted to bring your attention to Kopin's announcement from earlier today.

Kopin's new color VGA quality Cyber (indiscernible) systems have been integrated into a new, top-of-the-line digital still cameras from Konica Minolta.

Konica Minolta formerly introduced the camera late last week.

We were very excited about our partnership with them, and I will go into further details on the announcement shortly.

But first let's look at III-V product line.

Our III-V product revenues grew approximately 46 percent sequentially in the fourth quarter, fueled in part by strong sales of wireless handsets.

We also continue to experience stable demand for our band-gap transistors for wireless LAN and WiFi applications, a trend we believe will continue in 2004.

We recount (indiscernible) our development activities.

We continue to advance development of our GAIN-HBT.

GAIN is truly a second-generation transistor, offering reduced power consumption, improved filter (ph) quality and reliability while retaining transparency to our customers.

Our customer do not (ph) ever change or modify their manufacturing lines or processes.

So far, the results have been very impressive.

And we expect the transistor to be in design into power amplified circuits in the next quarter or two.

As I previously stated, we believe GAIN will deliver the performance, reliability, and power efficiency that manufacturers need to boost the performance of the next-generation wireless handsets and wireless LAN product.

In CyberLite, we're now shipping both half (ph) blue and green LED (indiscernible) tier one Asian customer.

They have another customer qualified.

LED (ph) are truly the next major III-V market for Kopin.

And we continue to invest heavily in the CyberLite to achieve even greater product performance.

Similar to our experience with CyberDisplay, we have used the further (ph) CyberLite to gain experience and traction with customers in great demand from Kopin's (indiscernible) technology.

We believe our next-generation products, which we expect to introduce later this year, will be in high demand among our tier one customers.

These included a half blue and green LEDs, as well as the 3-in-1 RGB chips (ph).

Turning to our CyberDisplay business -- we achieved a number of great accomplishments in quarter four.

Momentum is (ph) clearly as good as for Kopin's new color filtered CyberDisplay, a complete easy-to-integrate electronic viewfinder system called CyberEVF.

One of the designs wins we achieved in quarter four was with Konica Minolta.

As I mentioned, this design win is particularly exciting for Kopin, both in terms of the status of the customer as well as in the significance of the event.

Konica Minolta is one of the world's leading digital photography equipment manufacturers, and their development expertise is virtually unparalleled.

It is no surprise then that they select our color VGA quality CyberEVF as the viewfinder for their new 8-megapixel DSC.

With this new camera, both Konica Minolta and Kopin have made major breakthroughs.

This camera is the first commercially available digital still camera equipped with our VGA CyberDisplay.

The advantages of using our color display are described in detail in our release this morning.

But suffice it to say using our color display the image from the viewfinder is truly impressive.

Our alliance with Konica Minolta reflects Kopin's strategy of broadening its customer base from camcorder manufacturers.

In addition, as the first (indiscernible) application for one of our new color filter displays, this new camera will be the first significant color revenue generator for Kopin, an important step for meeting articles (ph) of generating roughly half of our CyberDisplay revenue in 2004 for color product sales.

For Konica Minolta, the company can lay claim to have the highest-resolution, most true-to-life viewfinder ever integrated into a digital still camera.

It goes without saying that we are looking forward to more collaborations with this industry leader.

Currently we're ramping our production for our color displays, and so far the ramp up has been smooth.

We are also experiencing an increase in the number of conversations we're having with military (ph) industrial partners, who see our ruggedized monochrome displays as the ideal for their special film and imaging applications.

Early this month, we announced that production order from FLIR Systems, based in Sweden, who will be use Kopin's ruggedized VGA quality CyberDisplay 640M module in night vision by Markalev (ph) for both Swedish and Norwegian armies.

Kopin worked with FLIR to specifically design a module for the company's forward (indiscernible) thermal imager.

This is actually a very important point.

As the applications using our CyberDisplay become more sophisticated, our customers are looking for Kopin to be more than simply a component supplier.

We are being relied on for our design expertise as well, as our engineers work closely with our customer to develop the optimum housing, optics, driving electronic configuration for their products.

This is the key part behind our development of CyberEVF viewfinder systems, and we believe Kopin's total system expertise is another barrier for entry for potential competition.

Konica Minolta and FLIR are just the beginning of what we believe will be a list of new customer announcement for Kopin this year.

Before turning to our outlook for 2004, I think it's important to put Kopin's 2003 accomplishments in perspective.

Last year, we embarked on an important initiative distributing Kopin's positioning in all RF3 (ph) products.

In our III-V product lines, we make advances in our development of GAIN-HBTs.

We also begin shipping blue and green CyberLite LEDs based on proprietary NanoPocket technology.

And CyberDisplay, which uses a total of 6 ultrasmall, high-resolution displays and viewfinder systems in the digital still camera, thermal imaging, and custom electronics (ph), including hand-model (ph) displays and personal entertainment systems.

We believe each of the activities (ph) set the stage for a very strong year in 2004.

Interesting, our color (indiscernible) display continue to grow, and the quality and time-to-market vary (ph) -- proposition of a complete CyberEVF viewfinder system is (indiscernible) above in the industry.

We anticipate this year of design wins in this year, and we believe shipments of our color displays will ramp steady throughout the year.

In III-V, we're moving forward with performance advances in both HBT and CyberLite products.

We foresee continued strength in our HBTs in wireless LAN applications, where we expect to introduce enhanced-performance LEDs later on this year.

As a result of our leadership in nanosemiconductor materials science, our products truly unique -- uniquely proficient in the intersection of mobile communications, computing, and personal entertainment.

Kopin's focus have always been this ultimate conversion of those three separate function -- applications.

Our HBTs are critical components in the wireless pipeline that fuels a new generation of advanced (indiscernible) and WiFi.

Our CyberDisplays and EVF systems act as a crystal clear video hub for the contents -- video, data, and graphics -- that flow across that pipeline.

And our CyberLite LEDs provide the portable solid-state lighting sources that complete the link between the user and the information.

Although the wireless marriage of voice, data, and video may not be perfect yet, a perfect union is not that far away.

Our proprietary technology, differentiable products, tier one customers, and strong financial resources put Kopin in the ideal situation sufficient (ph) to capitalize on those opportunities.

We would like now to open the call to your questions.

(OPERATOR INSTRUCTIONS) Blaine Carroll, Adams, Harkness & Hill.

It's actually Jed Dorsheimer for Blaine Carroll.

Congratulations on today's design win at Minolta.

That's really fantastic.

Thank you.

We're very excited about that.

You know, Minolta is very well known all over the world.

And their cameras are very high-end.

And it's a beautiful camera.

You should take a look at it.

John, you seem to have had a lot of design wins with the CyberDisplay.

And really, it seems like you're executing on your strategy to diversify into other markets.

Could you give us a little bit of color on what this -- from an end-market perspective, as far as percentage, what is coming from camcorder and what is coming from sort of other markets?

Yes, the question is really the ratio of camcorder business vis-a-vis the other applications.

Currently, of course, camcorders dominate our market right now.

And the Minolta camera is our first design win that was announced -- go into DSC.

Our application space, however, is much broader than that.

As we just talked about the convergence -- we're seeing really a lot of activities using CyberDisplay -- not just for future still camera, which we believe it is going to keep on -- would grow, because we have seen the middle-range digital still cameras are beginning to really seriously thinking about electronic viewfinders.

Remember, all consumer electronics product usually goes to the high-end first, and then goes to mid-end -- the mid-end guys want the same functions the high-end guys already have.

So I think the digital still camera will continue its growth throughout the year.

But we also see a lot of personal entertainment -- people using this display for personal entertainment purposes, which I think that we'll be very delighted to announce throughout the year.

And I think the ratio -- good question to answer the ratio is maybe by the end of this year -- how do we diversify from almost 100 percent camcorder to other applications.

So what would that percentage be?

By the end of this year?

Yes.

We will (indiscernible) 50 percent will be -- more than 50 percent be color displays.

Other than that, we should see how they come out.

Some of it depends on customers, because some of the personal entertainment systems -- both our customer and our sell (ph) we think will really sell like hotcakes by Christmas.

But you never know until it's done.

Sure.

So I guess with that being said and with the higher ASPs that your color products typically carry versus the monochrome, when you had mentioned steady growth for '04, is that -- does steady mean 10 percent, 15 percent?

What do you think we can see for growth out of your CyberDisplay market?

Rich, maybe you want to comment on that?

Our historical pattern -- and I don't think we're going to change it today -- is that we get one-quarter guidance on total revenues.

We don't break it out by -- guidance by individual product lines.

But I think we did announce (indiscernible) that this year would be a good year for us.

All right, but it would be fair to say that you should expect with all these new products beginning to ramp and the color replacing some of the monochrome as well as steady monochrome growth -- an ASP increase for '04?

Maybe, maybe not.

Another piece of the strategy, again, that we talked about is that we have a shrink display, a smaller one which comes in color.

And one of the beauties of that is we can actually offer that at a very comparable price to the existing monochrome, but because it's smaller, we actually get twice the number of die per wafer.

So even though the price might be the same, even a little bit less, it's color.

We actually get better gross margin.

So it really is going to depend on the sales mix between the high end and the low end.

(multiple speakers) Also to answer your question on (indiscernible) also -- is as we ramp up, the you (ph) is very important.

So far, we -- as I say, the ramp up has been pretty smooth.

But we should continue doing the spirit (indiscernible) we keep up a smooth ramp.

Great.

Just jumping over to III-V for a minute -- John, the CyberLite business -- could you give a little more detail or clarity on the next generation?

I know your strategy in the first generation was to come out with the lower forward voltage and not necessarily be the brightest or the lowest cost.

What does the second generation entail?

It's a very good question.

CyberLite was introduced last year with a lower voltage VF (ph) chips.

Two things that happened in the last year, of course -- one is, of course, we're aiming a key patent applications because of things happening in Asia, we thought the ASP was dropping too fast.

And secondly is that there are some people who didn't quite reach our voltage but they're getting closer to our low voltage.

So a few months ago, we announced we're going to develop a new set of products out.

These set of products -- suffice to say will not only combine the good features of previous -- last year, but (indiscernible) will make it much brighter.

And we're happy with our progress right now.

Great.

Did CyberLite -- did you break that out?

No, we have not -- I think we have not (multiple speakers) we have not broken it out.

Sure.

Would you comment on whether or not that was flat, grew, or declined, and by what amount?

No.

All right, let me ask you a different way -- your largest customer -- what percentage of the HBT revenue did they account for?

Well, the largest customer at III-V was Skyworks, and they were 21 percent of total revenues. (multiple speakers) And about 45 percent of total revenues were HBTs.

Tom Sepenzis, ThinkEquity.

Congratulations on the quarter and the guidance.

Just a couple of things.

One -- in terms of the III-V revenue, particularly wireless -- given the strong comments out of Skyworks and Anadigics and Tryclin (ph) regarding the March quarter, is there room for a little bit of upside if the turns business is strong?

Yes.

And can you talk a little bit about the ASP trends?

I think 15 to 20 percent annual declines is normal.

Is that what you're seeing now?

Or are you seeing a little bit better?

No, that's what we expect, and that's what we are forecasting for this year in our internal budget in units.

But you're dealing with an ASP of several hundred dollars.

And so the notes -- we're hoping for a number of (technical difficulty) it runs roughly about 2.5 million a quarter.

Okay, so that should stay the same going forward (multiple speakers) the expectation is that you'll remain cash-flow positive?

That's what we strive for.

Great.

Tom, actually (indiscernible) calm about this stuff (ph) FLIR Systems.

The thermal imaging system, as you well know, is actually (ph) continuously becoming more and more and more demanded by many militaries -- also semi-military outfits like home securities.

And our systems really are fully qualified and used by a lot of people now.

So we think this is -- actually could be a very good growth area for us.

And in terms of the LED business, you qualified a second customer.

Are they going to start ramping volumes in the March quarter?

Or is that still a ways out?

We certainly hope so -- (indiscernible) -- we're going to work on the details.

And I know you won't comment on whether it was up or down in December -- but March, what is your expectation?

Again, we only comment on total revenue.

I think the important thing -- we stated that as our strategy, is that we will keep our customers.

But meanwhile, we introduce new products.

And I think that's why -- we obviously put a lot of expectations on our new product development.

And we've been working on it for about five months.

And I think hopefully this will really make a difference.

Pierre Maccagno, Needham & Company.

Could you comment a little bit on the LED environment today -- the ASPs?

And how do see Taiwan competing in this market?

Which one -- on the LED?

Yes.

Okay.

Well, the LED on the keypad market (ph) -- the projects are stabilized.

The price -- there was a selling drop, a very steep drop actually -- the prices are stabilized.

So we don't see the severe drop that we saw middle of last year.

However, we are now focusing on different type of performance, and that price has never really dropped that much anyway.

So part of the reason is the Taiwanese are really not there in those higher-performance products.

It also seems to be that demands are going to higher, higher performance right now, which is also in sync with what we're doing now.

So the all-in-all is I think that the market is still growing.

The keypad business dropped and now is stabilized.

And the high-end products -- higher-performance product seems to be more and more in demand right now.

So are you planning to enter the color display -- the flash in the handsets?

Or what areas are you planning to enter?

For high-performance -- probably a lot of those -- all of the above.

Our customer package them, and it goes to different applications.

Earl Lum, CIBC World Markets.

John, with regards to the LEDs, the second customer that you've qualified -- is that for both colors, or just for one?

I am afraid you would ask the question.

We've got one fully qualified.

The second one is shortly (ph).

Okay, so you have one color qualified, and the second one is coming along.

Yes, coming along, yes.

And can you comment a little bit about what's going on with the automotive qualifications that you've been doing over the last couple of quarters?

And is there an eventual closure to that?

And when would you potentially expect that to be?

I think that's -- again, a very good question.

The qualification continues, but we think there are -- hopefully, our newer products would really change it -- make it move even faster.

Would you have to start a new qualification then with the newer product, or how does that work?

You probably know that more than I do.

It depends how the situation is.

If the product is really outstanding, usually people do a (indiscernible) -- mini-qual (ph).

It's actually try a (indiscernible) program for you.

If not, then they will do the whole, full-fledged qual.

But usually when you're already a customer to them, the qualifications usually go much faster. (multiple speakers) They (indiscernible) a lot of things are already done.

It's just (indiscernible) couple of things here.

All right.

If we shift over to CyberDisplay -- congratulations on the Minolta announcement.

Yes, we're very excited.

It's a great camera.

You should look at it.

The viewfinder really knock your head -- your feet off, because it looks like there's like a big TV in front of you.

If you look at the lineup that Minolta is introducing, certainly over in Europe and in Asia, because of the DiMAGE (ph) name that they're using for the trademark over there -- how would you expect things to progress as we move forward?

Is Minolta now a -- or has Kopin been selected by Minolta as a strategic partner going forward?

And do you anticipate that you'll have an additional 10 percent customer coming from the digital still camera market by the time we reach the end of 2004?

I think what happened was -- I'm going to digress little bit what your question is.

Your question is really how big is the Minolta going to be?

My answer is really, I think what they did is they put a (indiscernible) display, a VGA display into a viewfinder in the SLR type of cameras.

And that is -- that camera, actually, if you look at it is growing percentage-wise to DSC much faster.

If you want to know, DSC for the very low end actually go into the phones.

So normal (ph), now DSC is trying to replace photography camera, and I think it's a perfect one.

So what happens -- I don't believe Minolta will be the only one using our display in SLR cameras.

So (indiscernible) obviously additional people into it.

Okay.

So you're obviously targeting other tier one SLR camera manufacturers.

Or maybe they are targeting us?

Okay.

At this point, with 8-megapixel being pretty much towards the very high-end, would we expect then to see as other SLR camera suppliers look to use your product -- would it be in that type of category range, initially?

Yes -- I think -- like everything, the SLR camera is going to be make it look like -- feel like a full photography camera that (indiscernible) films.

And I think it's happening -- the 8-megapixel CCD is definitely going that way also.

What we're also seeing is -- remember, our product introduction for DSC is not just one high-end definite (ph) low VGA.

But we have 180 for the mid-end and 113 for the low-end.

And I think -- our goal is to -- hopefully to win all three ranges of DSC (indiscernible).

And John, just quickly -- what was the fab utilization at the end of Q3?

You said at the end of Q4, it got up to 70 percent.

Actually Earl, it was roughly the same.

Q2 was around 65 to 70, and here we are slightly -- a hair over 70.

Even with the big jump in the revenue side, the utilization didn't go up substantially?

Am I reading that correctly, Rich, or --?

Again, we calculated an overall companywide utilization.

Right -- that includes all three of the (multiple speakers)

Does not include CyberLite.

Does not include CyberLite, okay.

So it includes HBT and display.

And then just one final one to wrap it up, and I will let the next person ask some questions.

With regards to the wireless LAN, John, do you see that continuing to kind of range in this 10 percent level as we move forward?

Do you see as we move toward 802.11a that we might get an incremental bump from the wireless LAN Area?

Or can you comment a little bit about that?

Well, similarly, I think as you go from B to A, the content of using gallium arsenide versus (ph) silicon will increase.

As you well know, go back to (ph) wireless LAN -- obviously the ability to exactly how it's being used in it is still a guesswork, especially with big customers.

Big customers don't break down as easily.

The smaller customer we can tell, all right?

We know what they're selling to.

But guys -- our big customers are -- we have to really try to decipher the structure band-gaps -- could be used for phones or for wireless LAN.

And certainly we don't get the real breakdown.

But I do believe as we go from B to A, the content will increase.

We might get a bump -- but that we cannot integrate right now.

Sandy Harrison, Pacific Growth Equities.

Most of my questions, I guess, have been answered.

So I'll do a couple of housekeeping ones.

You know, I guess one of the questions with the strength we saw in handsets in Q4 -- and you guys definitely leveraged that case in your numbers.

We could see that.

Do you guys think you could have done a little -- a, I guess the first question is, did you see, or did your customers reflect to you, difficulty in getting components so that they could meet all of their demands?

And a follow-on to that would be, given that, do you think that if your customers could have gotten all the components they wanted, you might have been able to ship even more in the fourth quarter?

That's an interesting question, Dave.

Certainly -- with bits and pieces of -- these people think they can ship more if they can get everything lined up together.

And definitely a common come out (ph) left and right.

But how big is that to (indiscernible) from our customers is still a guesswork.

Got you.

As far as in the DSC market and in the Minolta announcement today, what do you guys think -- I mean, what's sort of your view, or what position do you take on what you think the size of the market and the opportunity is for the higher-end cameras?

And again, Rich, you had made some comments about some ASPs on some of the other products.

What are you looking at potentially for ASP for a Minolta kind of a design?

Well, the problem with giving you an ASP on Minolta is we only have Minolta.

People will get very upset, they would be very upset.

But I think I can answer your question this way.

SLR kind of camera -- digital camera -- the estimate we heard is around 2 million sets per year, so --

Got you.

So Minolta is, I guess, expected or pretty much under (multiple speakers)

Minolta (multiple speakers) this year, right?

And Minolta is a major player in that market.

Minolta is not really necessarily the major player.

There are a couple of guys equal or bigger than they are -- Canon and Nikon are bigger. (multiple speakers) So there are quite a lot of units there.

And Denmark is the fastest-growing market by percentage in DSC right now.

Got you.

Okay, and then -- it looks like we had a little spike in the R&D stuff or R&D expenses this quarter.

Does that -- Rich, do you expect that to continue at these levels, to come back and sort of what was behind it being up a little bit this quarter?

Again, we budget 15 to 20 percent.

We were up around 21.

And a lot of it had to do with some of the CyberLite that we'd been investing heavily in.

But again, I would suspect it's going to be in the 15 to 20 percent range.

We're very active on CyberLite the last four or five months.

Kalpesh Kapadia, Unterberg Towbin.

Hopefully there are more questions left to ask.

You always have good questions, so I am waiting.

Sure.

So CyberLite -- if you had only one customer, that means it was less than 10 percent of the total revenue for 2003, since it wasn't one of the 10 percent customers.

Is that the right way to read it, Rich?

That is correct.

And was it closer to 5 percent or closer to 10 for the full year?

I only need to disclose customers greater than 10, and I need to say no more.

Also, only one (ph) customer -- you don't want to review what is going on with that customer, anyway. (multiple speakers)

Fair enough.

You know, Matsushita dropped out of the 10 percent customer list for the year.

Is that because other customers were ramping larger?

Or is there something going on at Matsushita?

I think that -- what really happened is you saw the Samsung number had gone up significantly.

And I think it's pretty well known in the market that, particularly in the low-end products, Samsung is really taking share.

So it's a function of other customers growing fast.

Yes. (multiple speakers) The low-end camcorder is doing very well.

And Samsung is doing well.

Okay.

In terms of -- Rich, you mentioned that even though you may have (indiscernible) ASPs for color displays versus monochrome, the gross margin should be higher because you get twice the number of dies per wafer.

And if I look at your revenues from fabless (technical difficulty) 65 percent, and you expect 50 percent of that having color displays this year, what kind of gross margin improvement are we looking at from the 24 percent level that you reported?

Well, we would hope to exit this year in the mid-30s.

Exit this year in the mid-30s? (multiple speakers) So there should be a constant improvement through the year?

Yes.

And average should be closer to 30.

Yes.

Okay, the next question on -- you mentioned that there could be some upside to numbers considering what Skyworks (ph) and Triquon (ph) and (indiscernible) said about HBT if you get some turns business.

Are you 100-percent booked (ph) for flat to up guidance that you provided?

No.

As you know, or I should say (technical difficulty) as you know, the HBT business is very much a turns business, given the short cycle time that we use to manufacture the product.

And now the fact that we're carrying inventory -- quite frankly, the customer knows that there's (indiscernible) on the shelf if they need it.

So it makes it even more of a turns business than it was.

So we've got to make revenue every week.

The leadtimes are like -- shorter than two weeks, maybe?

As a result of the -- yes.

On housekeeping note, there was a nice positive change in working capital.

Is there something one-time going on there?

Or just a good working capital management?

It was working capital management.

As you can see, we popped up payables, and we were able to beat the guys up and get some cash in and so on and so forth.

John, a very high-level question.

Recently, Silicon Labs introduced a CMOS PA.

And as you know, your customers are more for gallium arsenide band-gap (ph) HBT.

How do you see the market receiving that product versus the GaAs products?

Yes, I think that question is really Silicon Labs introduced a PA in silicon CMOS.

And I am sure that you probably talked to many of our customers.

And so did we.

And the answer from my view is people are not worried, because as you well know, for power amps, they really need their high-speed, very low power consumption and you actually have high power and good reliability.

I don't think the part that you're talking about can satisfied that.

And meanwhile, with more and more staff going to handsets, the requirements even get tougher.

And that's why we introduce GAIN right now.

So I think stuff like germanium silicon a few years ago -- everybody thought germanium silicon HBT power amp would wipeout the gallium arsenide HBT.

It didn't happen.

And the noise level really comes down now.

So you're probably getting -- what do you think?

I mean, you obviously -- in this area, you know -- investigating more than we do.

So we don't think that is really a threat.

And I don't think our customers do so.

You can talk to our customers (indiscernible) comments.

Sure, next question on --

It's really a basic technical problems -- challenges that go for power amps.

And so far, there is really nothing like gallium arsenide HBTs.

Actually design rules are very different.

With (indiscernible) design rule, there's a silicon -- that chip is very large.

And also requires very fine design rules that (indiscernible) kill anybody who would want to do that anyway in the first place.

Next question, John, on the camera stuff -- DSC design win that you had with Minolta.

You mentioned a 2 million unit market opportunity for a high-end cameras. (multiple speakers) Does that mean that the color displays only targeted the high-end cameras?

Or it could even show up in a $300 3-megapixel (ph) camera?

Yes, I had to make it very clear.

The VGA display was targeted to high-end camera.

Although we never talked about ASP, obviously ASP is relatively good.

So it's probably not suited for the mid-end camera.

The middle-end camera would come out with 180k color, which is about 0.24-inch diagonal, versus the VGA's 0.44-inch diagonal, both (indiscernible) smallest display in the world.

But that was aiming at the mid-end camera, $300 to $500 digital still camera.

And then we came out with another one -- this is all just last year -- we came out with used the 0.16-inch diagonal, which is 113k display -- color also.

And that goes for a real low-end camera, which probably a $150 digital still camera.

So today we announced the high-end one, which is the VGA one, and which we now think that we hope and we expect that the middle-end and low-end camera is using the dollar to display (ph) -- hopefully will also -- a design win will be announced this year.

So by the end of this year, the 50 percent of your displays will be color.

And how many of them would go to or the low- to mid-end digital still camera?

It still early for us to say, especially -- you're talking about unit or volume of revenue.

Revenue -- obviously, the VGAs is a very lopsided.

They are much higher.

Also, we are now working on the camcorder customers, converting them from monochrome to color, also.

So there will be camcorder conversion, also.

So the 50 percent number is for unit volume or revenues of CyberDisplay?

I think it's revenue, right, Rich? (multiple speakers) More than 50 percent.

(OPERATOR INSTRUCTIONS) Blaine Carroll, Adams, Harkness & Hill.

I think all my questions have been answered.

It must have been a mistake.

Earl Lum, CIBC.

John, just quickly back to the CyberDisplay.

Nikon just introduced a camera that had a 0.44 diagonal display electronic viewfinder -- same size as yours, but yet it only offers a third less in terms of the pixel count at 235,000 pixels using a polysilicon TFT.

If you look at that kind of a camera -- typically, on the products that you're going to be designed into, is that a sole source situation, or would you potentially be able to share with another vendor if the performance was exactly the same?

Or how would that work as we move forward into looking at your design wins?

I think -- very good question.

The Nikon camera using a display which is actually -- I don't think it's that secret -- is a Sony-make display -- is a quarter VGA (ph).

So it's about a quarter number pixels as our view display.

Interesting thing is that as you can see, we make our display at 0.44, which is identical size.

So that allows replacement, displacement much easier for obvious reasons.

The difference is really, as you go into different zoom, you look at the picture -- you look at a Nikon camera, the viewfinder picture is very pixilated.

In many ways, it's very fuzzy.

And with the Minolta camera, the view looks very beautiful.

It's like a film light.

So I think -- of course to make that display a 0.44 (indiscernible) VGA -- nobody else can make it.

We're the only ones that came out with it.

As I say, there's a lot of dust in the market right now.

Certainly with Nikon's brand -- at this point, if they are now potentially fielding a camera that is in second place to the Minolta in terms of the resolution, one would expect that they would be looking at your technology.

Have you at this point -- can you comment on if you have sampled Nikon with your display?

We certainly can't do that because our customers would not be happy.

But we certainly talking to SLR -- all (ph) manufacturers right now.

Kalpesh Kapadia, Unterberg Towbin.

Just a quick question on guidance.

You said flat to slightly up sequentially.

And that comprises of CyberDisplay revenues growing from 12.4 million.

Does that mean that the HBT or III-V side would decline slightly seasonally?

Again, we only comment on the total revenue.

But you did say that CyberDisplay should grow sequentially.

If we did, I don't mean to have said that.

I don't recall saying that, to be honest with you.

So could you just clarify what would be the individual components of the guidance, or --

No.

As I said, we only give guidance on the total revenue line.

And on the gross margin front quarter to quarter, we should also expect the similar margins?

Yes.

I think Kalpesh -- I know what you're striving for.

I think that seasonally, obviously, you look at -- usually for consumer electronics, our first quarter usually tend to be the weakest -- also for display, also.

Of course, it's compensated by the fact that our current display goes to new models coming in.

So, I think it's not so clear that the III-V revenue comes down.

But we don't break it out, yes.

And at this time, we have no further questions in our queue.

I would like to turn the conference over to Dr. Fan for any additional or closing remarks.

Well, thank you very much for joining our conference call.

And we look forward to speaking to you again in the next quarter.

Talk to you.

Bye-bye.

That does conclude today's Kopin Corporation fourth-quarter and full-year 2003 financial results conference call.

You may disconnect at this time.

We do appreciate your participation.