庫力索法 (KLIC) 2017 Q3 法說會逐字稿

Greetings, and welcome to the Kulicke & Soffa Third Fiscal Quarter 2017 Results Conference Call.

(Operator Instructions) As a reminder, this conference is being recorded.

It is now my pleasure to introduce your host, Joseph Elgindy, Director of Investor Relations and Strategic Initiatives for Kulicke & Soffa.

Thank you, Mr. Elgindy.

You may now begin.

Thank you, Rob.

Welcome everyone to Kulicke & Soffa's Third Quarter Fiscal 2017 Conference Call.

Joining us on the call today is Fusen Chen, our President and Chief Executive Officer; and Jonathan Chou, our Executive Vice President and Chief Financial Officer.

For those of you who have not received a copy of today's results, the release as well as the latest investor presentation are both available in the Investor Relations section of our website at investor.kns.com.

In addition to historical statements, today's remarks will contain statements relating to future events and our future results.

These statements are forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995.

Our actual results and financial condition may differ materially from what is indicated in those forward-looking statements.

For a complete discussion of the risks associated with Kulicke & Soffa that could affect our future results and financial condition, please refer to our recent SEC filings, specifically the 10-K for the year ended October 1, 2016.

I would now like to turn the call over to Fusen Chen for the business overview.

Please go ahead, Fusen.

Thanks, Joe.

We're very pleased to reach the top end of our guidance with $243.9 million of revenue in the June quarter, generating $30.8 million of net income and $0.43 of EPS.

Excluding our tax credit and the noncash impairment charge, net income would have been approximately $45.1 million and $0.62 of EPS.

Our model continues to deliver strong operating leverage at these heightened levels of revenue.

As we continue to optimize our existing business, execute toward development goals and pursue and integrate new growth initiatives, we anticipate further enhancement to our model.

In addition to the higher level of business during the quarter, we were pleased to close on our latest acquisition of Liteq BV, which provides an additional vector of growth to our dedicated base of advanced packaging solutions.

During the quarter, we had also recognized revenue on our third APAMA offering, APAMA DA, which is geared toward advanced die attach applications.

I will share some additional information on these new opportunities after financial review.

During the June quarter, our ball bonder demand continued to stem from a diversified mix of end applications and customers.

As we have discussed in prior calls, the strength in this business is derived from 3 key areas.

First, digestion of copper capacity related to the effect of copper replacement cycle, which began around 7 years ago, are behind us, returning our business to a more normalized level of replacement demand.

Secondly, strong semiconductor unit growth combined with a healthy utilization is driving incremental capacity add.

Finally, wire bonding is commonly misperceived as a nonadvanced form of packaging.

A sizable portion of our current ball bonder achievements are supporting the latest generation of NAND and SiP applications, which are critical components to the fast-growing memory and mobile connectivity markets, more effectively, all globally produced NAND flash memory.

Wire bonding is a method utilized to connect each individual die within the stack.

For NAND, we believe we clearly have the leading position with over 90% market shares of die-to-die connections.

Similarly, for mobile SiP applications, wire bonding provides flexibility and cost advantage over Mass Reflow SiP, an additional segment we also provide dedicated AP equipment for.

Within this high performance wire bonding market, again, which support high volume production of advanced packages, we have a dominant share positions.

Our capillaries business provides additional insight of application and end markets served within the large installed base of wire bonder.

Based on this information, we believe approximately 20% of current wire bonding capacity serve advanced packaging applications such as stacked NAND and SiP.

Beside our advanced packaging opportunities, our wedge equipment business continued to remain robust and is driven by other interesting market trends.

The current level of demand continues to be supported by broadening growth prospect in automotive, industrial and power semiconductor applications.

Our average quarterly wedge equipment revenue fiscal to date was 30% higher than the preceding four year average.

This incremental demand is largely attributable to our execution in new nontraditional automotive wedge applications, which we anticipate will continue beyond 2018.

During the June quarter, we have also experienced ongoing progress within our volume base of dedicated advanced packaging solution.

Our dedicated advanced packaging solution, excluding wire bonding, accounted for $26.5 million during the June quarter.

We recognized revenue on 2 system sales of our advanced die attach platform, which marked the third product line of our modular APAMA tool.

This die attach platform targets advanced packaging opportunity in memory and image sensors.

Additionally, our Wafer Level Packaging solution, AT Premier Plus, has generated strong demand in the June quarter, supporting image sensor and also MEMS-based applications.

Considering our recent acquisition of Liteq, we now have over 6 dedicated advanced packaging solution derived from 4 unique platforms, which collectively target a $930 million market of dedicated advanced packaging opportunity by 2021.

We continue to be very well positioned as this opportunity begins shifting into a higher volume production.

I would now like to turn the call over to Jonathan Chou, who will cover this quarter's financial overview in greater detail.

Jonathan?

Thank you, Fusen.

I will compare the June quarter results to the March quarter.

Net revenue for the quarter was $243.9 million.

Healthy gross margins of 45.8% generated $111.7 million of gross profit.

Excluding our noncash impairment charge and favorable tax credit, we would have generated $45.1 million of net income and $0.62 of EPS.

As Fusen has discussed earlier, this strong performance highlights the relevance and profitability of our growing base of solutions.

Earlier in the quarter, we decided to take an impairment charge of $35.2 million related to the goodwill associated with our January 2015 acquisition of Assembleon.

The near-term outlook of this business was largely impacted by capacity digestion from strong fiscal year 2016 hybrid sales, but also the greater than anticipated reliance on wire bonding for high volume SiP applications.

Despite this noncash accounting charge, we continue to be committed to this business and platform.

During the quarter, we also resized a portion of our organization, which resulted in an additional $2.8 million of restructuring expenses.

Separately, we received a favorable tax credit of $20.9 million.

Excluding this favorable credit, we currently continue to target a long-term tax rate of 18%.

Turning to the balance sheet.

We ended the June quarter with a total cash and investment position of $593.9 million or $8.19 on a per share basis, up $19.7 million sequentially.

This excludes the Liteq BV acquisition, which was closed at the beginning of the current quarter.

On a book value per share basis, we ended the June quarter with $12.25.

Working capital, defined as accounts receivable plus inventory less accounts payable, increased by $65.5 million to $253.4 million.

From a DSO perspective, our days sales outstanding increased from 77 days to 79 days.

Our days sales of inventory increased from 81 days to 86 days, and days of accounts payable decreased from 66 days to 59 days.

Considering our recent acquisition, restructuring and current level of business, we want to provide some additional outlook to our September quarter operating expenses.

We are currently anticipating approximately $64 million of total operating expense in the September quarter, slightly higher than our model, due primarily to higher R&D expenses related to our recent acquisition.

We are in the process of completing our 2018 annual operating plan and will provide a more detailed outlook to our long-term model during our fourth quarter earnings call.

This concludes the financial review portion of our call.

I will now turn the discussion back over to Fusen for the September quarter's business outlook.

Thanks, Jonathan.

As disclosed in this morning's press release, we are targeting revenue to come in between $200 million to $215 million for the September quarter and a roughly $793 million to $808 million for the full fiscal year.

This represents an estimated 42% top line improvement over the same quarter last year and a 28% increase over the prior full fiscal year.

I wanted to spend a few minutes discussing the strategic intent and opportunities around the most recent acquisition of Liteq.

Liteq is our 6th product line in advanced packaging and serves lithography step for creating redistribution layer.

As a point of reference, redistribution layers are generally created prior to our traditional interconnect process.

We anticipate advanced packaging growth in Thermo-Compression, System-in-Package and Fan-Out Wafer Level packaging to require additional lithography processing, driving meaningful value creation.

A comprehensive understanding of both the bonding and lithography processes provides an additional value proposition to further attract broader optimization.

Liteq provide a unique solution and was the first lithography tool designed specifically for advanced packaging.

The significantly low cost of ownership and ability to handle wide variations of materials provides a disruptive approach to advanced packaging lithography.

We are very happy to welcome the highly effective Liteq team, who has demonstrated excellent execution by bringing these products through development and collectively add new insight and technical competencies to our growing organizations.

Relative to our big -- our other dedicated advanced packaging opportunities, we anticipate this new lithography business to expand our longer-term dedicated AP market opportunity by about 1/3 in 2021 from approximately $635 million prior to the acquisition to approximately $930 million in 2021.

This continues to be a long-term opportunity, as we are currently preparing the beta tool for qualification and have target several specific customers.

We anticipate the qualification process to take approximately 12 to 18 months.

We believe this is an extremely competitive solution with meaningful cost and throughput benefit over today's leading solution and capable of serving all advanced packaging technique.

We look forward to sharing additional progress with you on this new opportunity.

This concludes our prepared remarks.

Operator, we will now be happy to take questions.

(Operator Instructions) Our first question comes from the line of Tom Diffely with the D.A. Davidson.

So I guess, first on the advanced packaging market.

Did I hear correctly, so today, your served available market is about $630 million.

So sound like you had about $26 million, so you're about 5% of the market today?

I think, our available market is...

Yes, it's really more of a forward-looking 2021, we're looking at $930 million.

Okay.

Do you have a sense for how big you are in the marketplace today?

And what the projections are for growth over the next few years for your business?

Well, we think clearly, this is the -- one of the growth area, but it's only in the early innings of this -- in terms of that packaging side.

We do have some nearer-term kind of opportunity that we looked at, for example, in SiP and so forth.

So that's something that we're focusing very actively on.

Okay.

So how big do you think the market is today or this year, next year on more of a near-term basis?

I think, we're probably looking at a size about $300 million at this point in time in terms of what we're looking at and it's going to grow on a fairly more double-digit growth -- CAGR basis.

Yes.

Okay.

And then as far as the Liteq acquisition, so it's my understanding, this is a laser-based system that creates a system cost that's much lower than the competition right now, yet uses the flexibility of lasers to do a lot of different types of processing, is that correct?

Yes, that's correct.

Okay.

And then any update on where this tool is as far as, is it still in the R&D phase?

Is it at customers -- in customers' hands at this point?

I know it's pre-revenue, but I was kind of curious how far along it is?

I think this product is complete development and ready to ship to a customer for evaluations.

So in my remark, we target few specific customers, and we expect to ship 1 or 2 system by end of the year or beginning of next years, and we expect that evaluations should -- would take about a year or so time frame.

Okay.

Okay, great.

And then moving over to the core business.

You mentioned that the wedge bonder business is running about 30% above its kind of long-term average and you expect those trends to continue in 2018.

Curious if you can make a similar argument for the core wire bonding business that's obviously running a little stronger this year.

Do you see that continuing into next year as well?

Or is that a tougher market to call?

Well, I think the ball bonder, the market consist 2 part.

One is the replacement market and one is depend on the capacities.

So replacement, I think, has less to do with the capacity add.

I think in the middle of the year, it's difficult to forecast how good is going to be next year.

But based on our my current visit with customers, we believe next year can still be a good year.

Okay.

Yes, let me -- sorry.

If I can just add, we don't provide long-term guidance, but Semi just released some forecast data on the back-end of the equipment, and they're looking at 2017 spending up about 12%.

And they expect that to hold at the same level for next year.

Okay, great.

And then finally, you mentioned the NAND market as a kind of a new advanced market that's grown nicely recently.

Did you mention or say that you had 90% of the die-to-die connections in the NAND market?

Or was that a different comment?

Yes, I think that's correct, Tom.

Okay.

And so I assume with all the new capacity going into place, that should remain a pretty healthy market for the next few years?

Yes.

If we see the capacity for the NAND increase, we are confident -- our ball bonder, particularly on NAND, will do very well.

Next question is from the line of Krish Sankar with Bank of America.

I had a couple of questions.

Fusen, number one is, in terms of your wire bonder business, how much would you say today is NAND or NAND plus DRAM combined?

Is it in the 20%, 30% range, or higher, lower?

I think combined with NAND and SiP, it's also subject to quarter-to-quarter and year-to-year, right.

But I -- at this moment, I want a best guess is around 20%.

Got it, got it.

And is there a way you can quantify how much of wire bonders is going to IoT?

IoT take place in various form, so it's a little bit difficult to track.

But China is a big market of IoT and depend on how you define the IoT, but I would say, significant amount is for the IoT.

Got it, got it.

All right.

And then the other -- two other questions.

One is on the advanced packaging side.

Did you guys say that you had 2 system revenue from the Thermo-Compression bonder?

Is it right?

Or did I hear it wrong?

Our APAMA family originally have 2 platform, right?

One is the C2S and one is C2W and this was for Thermo-Compression Bonding, and we target for advanced memory and logic devices.

And we expect this market to be bigger for us next year and the next year beyond.

And the sub-product we just introduced today is called APAMA DA and this APAMA family is for the die bonding for advanced memory, right.

So I wish I cleared your questions.

Yes, yes.

Definitely.

And then just...

Go ahead, Krish.

Yes, okay.

And then also on the Thermo-Compression bonders, it looks like you guys are slowly gaining share.

But one of the pushbacks I heard is that, your solution is much more expensive than what's in the marketplace.

Is that true?

Or can you just highlight what the cost of ownership advantage TCB has over the incumbents today?

So TCB always more expensive compared to, say, other technology ball bonder and free chip because the throughput is a little bit slower.

And we are engaging with multiple customer in the qualification, and actually we see positive feedback from customers.

Hopefully, next year will be a bigger year for us.

Then the final question from my end on the Liteq acquisition, you guys did -- thanks for the confirmation that the product would be ready later this year, and in [qualification] with the customers.

Is ASML still involved?

Or did they did sell the stake to you guys?

So if I understand your question is, you are asking about ASML?

Yes.

Yes.

Is ASML still involved in the Liteq?

Or are they are out of it at this point?

Okay.

It's difficult for me to comment this question.

But they are part of the original investor helped to found -- start this company and still has some relationship with the company.

Our next question comes from the line of Craig Ellis with B. Riley.

Congratulations on the nice execution, guys.

I wanted to start off following up on some of the NAND comments.

Fusen, as we look ahead and if the 3D NAND players continue to do what they've been doing with 3D and what they did with planar, which is increased bit density and produced bit cost every year and give us an increased number of layers every year, so from this year 64 to 96 next year.

How do we think about the growth of your NAND business?

Does it really -- does it grow with bit output?

Or does it depend more on the annual increase in layer count for 3D NAND?

So actually all wire bonding work for both 3D NAND and such stacking NAND.

So I would think this is probably more related to a bit output, because we cover both the market.

That's my short answer.

You bet.

And then you've talked in the past about services.

I may have missed it, but I didn't hear you -- I don't think I heard you saying anything about services.

But with an installed base of 155,000 devices, can you give us an update on how you're thinking about the services opportunity and steps to take services revenue from 20% to 30% of mix?

Okay.

Actually, recently, we just formed a group try to grow the service spares, refurbish and consumable.

And this group just formed with very clear direction.

So I think start from 2018, this would be my -- would be our focus.

And most of the company, I think I have a model.

The post sale should be about 30% of total revenue, so that would be our goal.

Currently, we are around 20%.

And it would not be totally recognized in 1 or 2 years.

So hopefully in 3 or 4 years, we will try to reach that goal, 30% of our total revenue.

And then lastly from me and this may be more for Jonathan.

Jonathan, with operating expense going up to $64 million in the September quarter, the variance versus where it's been running in the 50s, is that all Liteq?

Or are there investments that the company is engaged in related to new products either in the traditional portfolio or with advanced packaging that would account for the increase to $64 million?

Yes.

I think there is -- clearly there's going to be some additional expenses coming in basically from Liteq, because of the fact that we are incubating a pre-revenue company.

And -- but we also are taking out some cost in the company as well which has results on operating expense.

So for this current quarter, it's similar to what we just reported -- what I just guided.

However, since we're in the process of actually doing our annual operating plan, I would say, we'll basically have a better kind of clarity and to kind of share with you during next fiscal year, our Q4 earnings call.

But just in terms of, just looking at where we are today, where we've been guiding in terms of the modeling on the OpEx side, $48 million plus -- basically 4% to 6%.

It is still the right level.

I would say, it might be slightly up from there after the budgeting process.

But we do look at our operating expenses very closely, our cost structure and we look at our breakeven revenue, which we guided in the past, and we'll continue to do that going forward.

That's helpful.

And then just a follow-up.

Since you mentioned that you are in the annual operating review.

This year versus a year ago, 2 big changes, new CEO with clear vision on where to take the company going forward, and secondly a significantly increased amount of EBIT that's flowing to the model due to a healthier market and very good execution by the KLIC team.

With respect to using the annual EBIT or the cash balance beyond strategic deals like Liteq, what should investors expect as we look out over the next 6 to 12 months?

I guess this is a -- in terms of -- in short, it's really about use of cash, capital allocation.

I think the way to our approach in terms of the prioritization of actually use of capital, is still in line with what we have done in the past in terms of basically organic -- basically investments.

Acquisitions basically expanding our portfolio like Liteq and looking at other opportunity that can actually allow us to fill out our portfolio of offerings to really better to serve our customers.

And basically the third priority would be basically return of capital to our investors through some basic repurchase program.

And as mentioned in my earlier call, we still have some remaining, basically a portion of $4 million, $5 million left on the current repurchase program, which will expire around August 14 of this year, and we'll share more news with you after that if -- when that runs out.

Our next question comes from the line of David Duley with Steelhead Securities.

Regarding the -- I think you mentioned you recognized revenue on a couple of APAMA tools during the quarter.

Could you talk about what the application was for those tools?

Was it die-to-die attach?

Or what were the applications?

CMOS and image sensor.

Say that again?

CIS.

CMOS -- basically a -- a image sensor.

So Dave as I mentioned, our APAMA family, there are 3 products.

We already introduced 2 product in the market, these are Chip-to-Substrate and Chip-to-Wafer.

This is TCB.

And we expect TCB to grow bigger next year and beyond.

And the product we introduced today is a die attach, is still in APAMA family, and focus in the memory and the CMOS imaging sensor market.

And we recognized 2 revenue in this quarters.

Okay.

And TSMC has been talking about, I think, a new packaging technology called CoWoS, I believe.

I can't remember the exact acronym.

But it sounded a lot like they were moving to die-to-die attach or wafer-to-wafer attach.

Is this something that the APAMA can address?

Yes, we believe so.

It's something that -- yes, that could actually use the Thermo-Compression Bonder technique.

So would -- I guess, maybe just elaborate on that a little bit.

Typically when TSMC moves into new packaging technology like InFO, they end up spending a lot of money.

So I'm just wondering is TSMC talking about this new packaging technology a major opportunity for the company going forward?

Or is this something that might be 1 year or 2 out?

So Dave, in short summary, our tool are capable to do a pick and place and Thermo-Compression, and we are engaging multiple customer, almost all the key player in advanced packaging.

So you name specific customer, and the short answer is, our tools are capable to be in this market.

Okay.

And then you talked a little bit about the copper wire bonding replacement cycle.

Could you just elaborate what you're seeing there with the installed base of copper wire bonders.

Is there a certain pad pitch where they're having to upgrade and you're seeing more of an upgrade cycle now than you have in the past?

Is there something different now than there has been in the last couple of years as to why that business is so strong?

Well I think -- let me answer that.

I think the replacement kind of the model that we've shared in the past is still pretty -- it's in line with what we have shared in the past, that our team, in terms of how we suggest how to model it for the investor.

So 150,000 of total wire bonding in field.

There is approximately 90,000 that -- basically that can be basically replaced.

The replacement cycle is 10 years and about 9,000 per year in terms of the annual replacement of that once it's -- obviously, other customers would depreciate that faster.

But we look at about 9,000 per -- and then there is incremental basically capacity increases, which is another 1/3 of that.

So that's really the model chalked out.

Okay, last question from me.

Did you have any 10% customers during the quarter.

If so, what were their percentages?

No.

Our next question is from the line of Manas Tiwari with Value Investment Principals.

I have couple of questions.

So my first question is regarding the recurring revenue that you target to -- from current 20%, you guided to make it around 30% of sales.

So what would be the time frame to achieve that target?

If I -- you broke up a little bit.

If I recall -- if I kind of rephrase your question it is about service, right?

Service moving from 20% to 30%, is that -- and what is the time frame?

Is that...

Yes, yes.

Okay.

I believe -- basically, let me just kind of reiterate what Fusen said.

It's going to take some time, but that is the sort of the industry target that we're trying to get to, in terms of 30%.

We're currently at 20%.

It's going to take a few years to get there.

But we have actually put in place from an organizational design perspective to put some focus and emphasis to grow that part.

Okay.

And what is your current market share in the LED segment?

And where it -- how much do you see this market to grow in -- say in couple of years?

At this point in time, obviously, LED market is quite large, and we basically have about 14% LED sales in terms of our current quarter -- current reported quarter revenue.

So that's really -- it's already up from the previously -- if you recall, we generally used to basically have only 5% on the average on the LED side.

So we were actually growing in that piece, then this is just a ball bonder LED related sales.

Okay.

That's helpful.

And the last question from my side.

What is the equipment utilization rate of your customers?

Because the last time on the call, I think you mentioned it was in high-70s.

Has it crossed 80%?

I'm so sorry.

You broke up.

Can you repeat that and maybe speak a little slower?

What is the equipment utilization rate for customer?

Because the last time I remember you said it was in the high-70s.

So is it still there?

Or has it crossed 80%?

Yes.

So it's about the utilization rate, where is it at?

We're -- in the past 2 quarters, about 78% and hasn't crossed over 80%.

But as mentioned earlier, if it crosses 80%, there's going to be some investment to add capacity for most of our customers.

Okay.

And when do you expect that to happen?

It's hard to say.

This is our customers.

We track their data.

I think right now it's to a positive curve.

It's sort of flattish and positive.

So I am not -- it's hard to predict when it's going to cross over.

(Operator Instructions) Our next question comes from the line of Tom Diffely with D.A. Davidson.

Just a quick follow-up on the memory market.

I think you mentioned that, it was roughly 20% of your business today and you had pretty high share -- very high share in it.

And I'm just kind of curious, it seems like in the past Asia-Pacific was a big player here.

So I'm curious what's changed in terms of either your products or the technology that give you such a nice market share today?

So Tom, more specifically, I think I make a comment.

The question I answer, roughly 20% of our ball bonder, of course, this has a lot of variation quarter-over-quarter, year-to-year.

Our support in 3D NAND and also SiP, so that's a longer explanation.

The reason, I think, we are high in the stack of NAND is, because our intrinsic capability.

We have a special looping capability that we can do a looping for 3D, right.

And we are also very experienced in this ball bonder business.

So that enable us to occupy roughly 90% of the market share for the NAND.

And SiP, I think wire bonded SiP is still the most effective -- cost-effective way to do SiP.

In fact, in some customer we took market shares, from -- they come from a flip chip to SiP ball bonder.

(inaudible)

Okay, great.

All right.

And then just maybe 1 more quick follow-up then on the gold -- or the copper replacement cycle.

Is it that the tools are just too slow?

Do they wear out?

What is driving actual replacement?

Well, I think there are many reasons, right?

One is the throughput and one is capability.

And of course, also, when too old, also has some reliability.

So we estimate roughly yearly the market size about 9,000 system per year.

And with no more semiconductor growth rate, it's roughly -- assume it's about 3% to 6% growth rate roughly.

I think the market size for capacity add is about 3,000 systems.

So that's our estimation.

Yes, I think if I could add.

In addition to just straight replacement or a capacity add, there -- just to -- for your information, the ball bonder that we just sold in terms of third quarter is 86% copper capable.

So there is still a little bit of actually conversion going on, even though it's -- we have actually said that it should be probably 75-25 kind of mix eventually.

We're probably at 70% at this point in time.

So copper cable -- and also I would say the performance.

The latest ball bonder is going to perform a lot better than even about 3, 4 years ago as I mentioned in earlier calls, to perform the kind of -- the looping process that we -- that some of the customers and package will require.

The next question comes from the line of Edwin Mok with Needham & Company.

The first I have is, die attach tool that you guys talked about, what's the advantage of your product over your competition?

We have a bit of higher accuracy.

And so we in the case of tool market only in few months, we already captured market shares.

So I think that accuracy probably is the biggest development.

Accuracy.

Okay great, that's helpful.

And then -- so recently we -- there is more NAND guys come out and talk of all these kind of newer package that they're trying to introduce that is using through-silicon via.

Our understanding is that's for a high-end data center application.

Have you seen the NAND markets starting to adopt through-silicon via?

And if so, how are you guys positioned in that?

Well, Edwin, for the TSV, this actually is not very a straightforward process or not as easy as a ball bonder.

You need to have a lithography step.

You got to have a higher experience in [couple of reading] and higher defect.

So I think at this moment, all we are seeing is the majority handled by our wire bonder.

And, of course, there will be some segment for TSV.

But I think at this moment, the volume is not compatible at this moment.

I see.

So you believe it's a very small part of the market.

Are you guys positioned for those opportunity though if they decide to produce some chips on TSV?

At this moment, we do not have a deposition.

We do not have a CMP.

We do not have Etch.

So at this moment, probably a little bit difficult for us.

But wherever TSV is part of a vertical interconnect like a copper pillar, we actually have a tool called AT Premier, is to create a vertical interconnect like a copper pillars.

And we are trying to look for more and more market for our AT Premiers.

And also I think our TCB, I think, is also being -- okay, after customer finish a TSV, our TCB is also capable to serve in this market.

Our next question is from the line of Steven Pelayo with HSBC.

Just wanted to follow-up on LED there.

A lot of talk to some very aggressive front-end MOCVD capacity going into China over the next year or so here.

And it sounds like it's been strong for you guys at 14% of the June quarter revenues.

I know it also sounds like you are going to be focusing more in this area.

So I guess, I'm wondering, two different questions.

Can you characterize kind of your breadth there?

Is this why customer breadth, why geographic breadth there in LED?

How are you particularly differentiated in LED in that environment?

And then, can I ask you to kind of put on your forecasting cap and how does LED look in the second half of this calendar year?

Does it sustain as mid-teens kind of run rate?

And what you think it ultimately is as a percentage of revenues going into -- in 2018?

Yes, Steve, maybe I could take a shot at answering your question.

I think historically, we were very selective or highly selective in terms of the LED market, because we do have the best performance basically bonder out there.

And we had a -- I'd like, to say, we had a bit of price premium over our competitors out there.

And since this market is really -- it's available and there is some emerging customers that we liked.

We basically have actually increased our focus in that area.

Using the platforms that we currently have and we look at this as all fixed costs were in there, for ball bonder, this is all contributing margins, it flows -- falls through very nicely.

And this is what I said about -- it's a price-sensitive market and that the gross margin for our business for that particular, those sales will be a little lower and which may have some impact to our overall gross margin, which we did allude in the past that, we're okay if it slips slightly below 45%, but we've been consistently above 45% for the company.

So we are looking at this a little bit more strategically in terms of how we actually play in that space, working with key customers that are emerging in terms of that space.

Let me answer a little bit differently.

I think you are asking about the whole LED industry.

The LED we mentioned we participate actually is in a ball bonder.

We don't have MOCVD, we don't have other tools.

So particularly I think at K&S, we are focused on the ball bonder.

And historically, we was not very aggressive in this market.

But since last year, I think, we put more effort, and we are seeing increased demand and increased business for LED.

I wish I answered your questions.

Yes.

I'm very aware that you guys focus on the assembly side of it all.

I'm just talking about there seems to be a lot front-end capacity planned for LED over the next year, lot of MOCVD reactors shipping.

Those eventually will be units that need to be packaged.

And so I guess, I'm trying to understand the breadth of your LED customer reach today and what does that mean for next year?

Do you think it's going to be mid-teens percentage of calendar 2018 revenues for you?

And even in the second half of the year do you sustain kind of this mid-teens percent of revenue too?

Well, I think we have -- we believe we can increase the market share in LED.

So we currently have a low market share, say 10%, 15% -- 15%.

So to double that I think in 2 years, I think is doable for us, just in our opinion

(Operator Instructions) Thank you.

At this time, I'll turn the floor back to management for closing remarks.

Thanks, Rob.

Before we close, I wanted to inform investors that we will be participating in D.A. Davidson's 9th Annual Technology Forum in New York City on August 15th.

Additional details on past and future events are available at investor.kns.com.

Thank you all for the time.

As always, feel free to follow up directly with any additional questions.

Rob, this concludes our call.

Good day.

Thank you.

Thank you for your participation, ladies and gentlemen.

You may disconnect your lines, and have a wonderful day.