FormFactor Inc (FORM) 2007 Q2 法說會逐字稿

Good day ladies and gentlemen and welcome to the Second Quarter 2007 FormFactor Earnings Conference Call. My name is Jeremy and I will be your coordinator for today. (OPERATOR INSTRUCTIONS).

I would now like to turn the call over to your host, Ms. Brooke Deterline, Vice President of Investor Relations. You may proceed.

Brooke Deterline. Good afternoon and thank you for joining FormFactor's Second Quarter 2007 Earnings Conference Call. With me are Igor Khandros, Chief Executive Officer and Ron Foster, Chief Financial Officer. Igor will provide a summary of our second quarter performance, review our market segments and provide an update on outlook and long-term strategy. Ron will then take us through the financial operational details and provide guidance. Please note during the third quarter of 2007 the Company will be presenting at the Pacific Crest Technology Leadership Forum in Vail, Colorado on August 7th. As other events and details become available, we will make additional announcements. Finally, before I hand the call over to Igor, I will review our Safe Harbor statement.

During the course of this conference call we will make projections within the meaning of the Federal Securities law including statements regarding FormFactor's growth and financial performance as well as our strategic and operational plans. The forward-looking statements are based on current information and expectations and are inherently subject to change. Actual results may differ materially and adversely to those in our forward-looking statements due to various factors including, but not limited to, the rate at which customers adopt the Company's newly released architectures, technologies and products, the rate at which customers implement manufacturing capability changes, make the transitions to smaller, nanometer technology nodes and implement tooling cycles, the Company's ability to effectively drive it's product development plans, introduce new products that meet its customer's testing requirements, the Company's ability to continue to ramp design and manufacturing capacity and increase efficiencies in its operations and the Company's ability to obtain tax and other cost advantages from its expansion of manufacturing operations in Singapore.

Please refer to the Company's recent filings on Form 10K and 10Q for more detailed information regarding the relevant risks and uncertainties. FormFactor undertakes no obligation to review or update any forward-looking statements or update the reasons actual results could differ materially from those anticipated in forward-looking statements.

Finally, a breakdown of revenues by market and geography and a GAAP to non-GAAP reconciling income statement are available on our website as is schedule reconciling our GAAP and certain non-GAAP financial guidance with respect to FAS 123R stock compensation expensing.

I would now like to turn the call over to Igor Khandros.

Thank you, Brooke. FormFactor delivered another record quarter with revenue bookings and operating performance all reaching historic highs. Advanced protocol demand grew yet again this quarter as multiple market drivers across segments propelled the market's growth. Revenues increased 12% sequentially to $114.1 million, non-GAAP gross margins were 54% and operating margins were 25.8%.

We improved our production performance by reducing lead times and continuing to invest resources in new technology and product development. Market demand for DRAM remains strong in the second quarter driven by DRAM bit growth, the ongoing transition to 70 nanometers and 1 gigabit driven by Vista and seasonal strength in the mobile business.

(inaudible) accelerated through the first half of the year. We believe the transition to 70 nanometers is proceeding as expected. Typically a two-year event, this transition is still in the early stages with the majority of the opportunities still ahead of us. We expect to see increase in 70 nanometer design for the (inaudible) of 512 megabits, DDR2, 1 gigabit DDR2, DDR3, mobile and graphic DRAM. The growing memory requirements for Vista and Microsoft Office 2007 will also move the market towards 2 gigabytes of DRAM content for PC throughout 2008 contributing to the strong DRAM bit growth of greater than 80% in 2007.

As manufacturers accelerate their migration to 70 nanometer in the second half to support the DRAM bit growth and improve the profitability of persistently low DRAM prices, this should translate to continuing DRAM probe card growth for the rest of 2007.

FormFactor will continue to lead the DRAM market as our latest product platforms, PH150XP and Harmony XP, combined to address the current and future test needs for mobile commodities and graphics DRAM. PH150XP solves the challenges of very high thin density on smaller devices and is the most cost effective solution for smaller devices with higher die count per wafer. With over 90% of the market currently in this range of 512 megabit density and [glow], our PH150XP product meets today's testing requirements in its four touchdown 300 mm capability to test 256 devices in parallel.

Our latest Harmony XP product is a full wafer contact focusing on larger device sizes. Harmony XP facilitate the optimal performance from customers' entire test cells in two to three touchdown 300 mm testing of 1 gigabit or higher density devices. It increases their throughput and tests their (inaudible) resulting in higher productivity and lower cost of tests.

This FormFactor's proprietary TRE (Tester Resource Extension Technology), the Harmony XP platform can be expanded to test more devices in parallel and accelerate the move to one touchdown in DRAM wafer testing in the near future. Harmony XP will also enable customers' road maps by scaling to and supporting (inaudible) pitch, the smaller step size, highest probe density and highest probe columns in excess of 50,000 pins. We will save in Harmony XP as customers make the transition to 1 gigabit density over the next year and our product will enter the market in volume by year-end.

Our known good die business performed very well this quarter driven by seasonal strength in our modal RAM and NOR businesses, service and consumer and modal and applications. Adoption of mobile known good die application drove strong high frequency tests of probe business while the transition to 70 nanometer DDR2 resulted in increased wafer level burning business. With FormFactor's known good die products now in volume production at multiple memory suppliers and our known good die revenues up over 60% year-over-year, we are seeing validation of our key [macrotesis] for the test market that more testing is moving from package to the wafer. With the increase in packaging cost, more customers are searching for ways to cut costs, especially with the current commodity DRAM pricing environment. By performing more validation at the wafer, package yields are improved and the wise performance is proven early at making the benefits of wafer level testing over package testing clear and further driving the migration to wafer tests.

We expect known good die products to perform well in the third quarter consistent with the seasonal consumer global cycle and continued strength in the transition to 70 nanometers. So in the long term we anticipate continued adoption of FormFactor's known good die products to be a major contributor to our growth. FormFactor is the only probe card supplier that offers a complete suite of known good die products providing unique benefits of higher parallelism, higher temperature testing and high speed testing to customers who are uniquely capable of introducing new generations of known good die products for future higher performance devices which will be increasingly sensitive to electrical signal integrity requiring our differentiating KGD solutions.

Our Flash business had a strong second quarter driven mostly by known good die for NOR and specialty NAND product. We continue to see strength in the segment as most customers move towards higher parallelism to decrease their costs of test and go with high-speed testing to enable known good die application.

We made progress in the development and qualification of our NAND Harmony OneTouch product this quarter and recently demonstrated superior performance on our Harmony OneTouch probe cards in production at a customer. These cards reduced their damage, improved electrical reliability and provided higher up-time versus other full wafer contacted technologies. We have also received new design wins for this product.

Our overall design capacity has been constrained due to an increase in new design demand and high design intensity associated with new product platforms and our Harmony learning on various test platforms has continued. Though we did not expect to see a significant increase in NAND Flash revenue in the third quarter, given customer device two-end cycles and design constraints, we plan to focus our efforts on improving design capacity and lead times and increasing our Harmony penetration in the NAND market.

Our logic SOC business continued to perform well driven by the new technology node transition for area array flip-chip microprocessor products. Development of our new products with a peripheral [pin-out] wire bonded device market continued this quarter as we introduced our first fine-pitch high parallelism product for automotive applications.

We completed the successful qualification of this product at our first early adaptive customer and anticipate multiple design wins over the next several quarters with volume production in early 2008. This product provides a step function increase in parallelism and addresses the many unique testing challenges associated with the (inaudible) environment including increased test times due to embedded memory, high temperatures and probe test page and size reduction.

Similar to our answering to the DRAM market years ago, we are well positioned with our advanced MEMS probing technology to redefine SOC market which is currently testing only one or two devices at a time. As complexity and testing cost and requirements increase, FormFactor will lead the logic SOC market to lower overall costs of tests through higher parallelism.

We continue to see the wire-bond market as mainly a 2008 opportunity recognizing that this is very different from our distant markets in that both customers and devices are very diverse. Thus, we anticipate customer acquisition, qualification and volume ramp to occur over a longer time frame first with design wins and then increases in both our number of customers and probe card volume for design.

Our overall outlook for 2007 market for advanced wafer probe cards remains unchanged from the view we provided in January. The advanced probe card market remains on track to grow with a robust 25% in 2007 compared to roughly flat growth expected for the entire semiconductor [category] an industry. FormFactor is well positioned based on the strength of the DRAM market and increasingly healthy NOR Flash market, continued emergence of the NAND specialty market and the strength of our existing Logic SOC business.

Similar to the trend last year, we expect to see fourth quarter seasonality in the consumer segment potentially upset by the accelerated migration to 70 nanometers and new market penetration. Our deep portfolio of industry-leading technologies and products, execution of the increases in design capacity and successful roll-out of our new products should drive our near-term success.

Now I'd like to take a moment to discuss the competitive landscape as we see it. Advanced wafer probing has been gaining importance for semiconductor manufacturers leading to steady growth in the advanced probe card market. This trend has obviously attracted an increasing number of companies to this market, some from the Legacy side of tungsten and vertical needle manual cards and from 2D and 3D MEMS.

FormFactor's 3D MEMS probe cards with wafer processing based, lithographic precision provide superior probe and accuracy, manufacture ability and scalability. Touch scalability is crucial to enabling customers' next generation technologies and performance requirements as well as to reducing their overall cost of tests. We believe FormFactor's proven proprietary with 3D MEMS technology invented and (inaudible) by FormFactor will continue to lead the probe card market growth. While other technologies will co-exist, we believe they will have relatively limited growth opportunities. FormFactor is the number one player with over 50% of the overall advanced wafer probe card market and we generate nearly 5 times the revenue of our two nearest competitors.

Additionally, we are the largest advanced probe card supplier in every geographic region including Japan. Our core strategic relationship with our customers worldwide, as evidenced by our recently announced agreement with Elpida, will enable us to integrate and align long-term test strategies to ensure consistent test cost reduction year after year. The combined strength of our technology leadership, our comprehensive product portfolio, strategic customer relationships and global manufacturing capability including our extension into Singapore differentiates FormFactor from the rest of the industry.

Historically we've witnessed an increasing competitive activity with every shift in test and parallelism. We observe that in the previous shift from 9 touchdowns to 4 touchdowns and we expect it again with the current shift from 4 touchdowns to full wafer conductors in DRAM. We are confident that FormFactor is as uniquely positioned to enable and define the industry's transition to 1 touchdown DRAM wafer probing as we were in defining the 4 touchdown transition. To give an example of one set of the future challenges that we plan to solve, DDR3 will drive about 20% increase in the number of pins per die over the year or two. This will potentially translate into up to 100,000 pins per probe card, one touchdown way for contactors, when DDR3 is ramping into late 2008. We also anticipate DDR2 contactors with over 70,000 pins during this timeframe. FormFactor 3D MEMS technology and our world class MEMS R&D and HVM factory are best positioned to meet the challenges versus any competitive approach.

Another example of our technology leadership is our recent demonstration of a new breakthrough MEMS spring contact technology capable of full area array, 20 micron test pitch. This contact technology is well suited to address the future testing needs of all semiconductor devices. We will continue to create new markets and new growth and these new capabilities are and will be protected by our strong IP.

Today, FormFactor is one of several companies leading important market segments related to semiconductors. But what makes our position unique in today's semiconductor field is that we're leading the growth market during the times when profitable growth is hard to come by in our industry. This uniqueness presents a complexity in interpreting FormFactor's competitive landscape from an investor standpoint.

The important point is that the competitive dynamics in our probe card market are not a zero sum dollar gain where one company's gain is automatically another company's loss. In our growth market, all of the growth dollars cannot be consumed by one company due to the custom nature of probe card products. But, as in the past, we plan to continue to redefine wafer test economics and we will continue to amplify advanced probe card markets. Naturally, competitors will have the opportunity to participate in the growing market but as the catalyst for transformation and the growth, FormFactor is well positioned to continue to be the long-term and primary beneficiary and to outgrow the market.

In summary, the second quarter of 2007 was another strong quarter for FormFactor. We are very pleased with the Company's achievement of record revenue and profitability, expand the factory capability and continued new product development. I would like to thank the entire FormFactor team for their hard work and efforts in making this another successful quarter.

I will now turn the call over to our CFO, Ron Foster, who will elaborate on operation, operating results, financial performance for the second quarter and will provide guidance for the third quarter.

Thank you, Igor. Revenues for the second quarter increased for the tenth consecutive time and reached record levels at $114 million, up 12% over the first quarter of 2007 and 24% over the second quarter of 2006. DRAM revenues increased 6% sequentially to $80.1 million and accounted for 70% of total revenues in the second quarter. Flash revenues were healthy and increased 21% from the first quarter to $20.2 million and 15% over the second quarter of 2006 representing 18% of total revenues in the second quarter.

Our strong performance in Flash this quarter resulted from healthy growth in NOR due to a significant customer's high volume ramp. While we did receive some revenue from our NAND Harmony OneTouch product in the quarter, overall NAND revenue declined. Production shipments and qualification at additional sites continue with one customer while qualification efforts with additional customers have been slower than expected. The differentiating capability of our Harmony platform has been demonstrated in production and qualification trials and we remain confident that we can penetrate additional customers as we progress in our qualification activities.

Logic SOC revenue was $13.8 million in the second quarter comprising 12% of revenues. This represented an increase of 37% sequentially and 82% over the same period last year. The flip-chip Logic business was very strong in the second quarter driven by the ongoing transition of one key customer's MPU's and chipsets to advance technology nodes.

Revenues from known good die products consisting of wafer level burning and HFTAP were strong at $17.6 million, an increase of 41% over the first quarter and 62% over the same period last year. This growth mirrored the strength in our NOR and mobile RAM businesses resulting from the seasonal increase in demand for consumer devices and from increasing adoption of known good die and new applications. The breakdown of revenues by market and geography is available on our website.

Bookings for the second quarter grew to $130.3 million with increases in all segments. Increasing adoption of our new advanced product platforms and improved market visibility by our customers drove strong bookings as well as turns reduction from 60% in Q1 to 50% in Q2. The order file continues to contain an unprecedented level of new design demand as it has throughout the first half. This is a positive indicator of future business opportunity but is stretching our design resources even with the additional hiring we have executed this year.

Now I would like to highlight some of the key GAAP results. Non-GAAP results will be called out where it provides additional value. A schedule of that provides GAAP to non-GAAP reconciliations is available on the investor portion of our website.

Gross margin for the quarter was 53%, a 500 basis point increase from Q1. Stock comp expense declined 700 basis points quarter-over-quarter. Non-GAAP gross margin of 54.0% was down slightly from 54.2% last quarter and in line with our target range of 53% to 55% for the fifth consecutive quarter. This is noteworthy as we started production of key new technologies in the factory during the quarter and began investing in Singapore manufacturing resources as well.

Operating expenses for the quarter were $37.4 million, or 32.8% of revenue compared to $37 million or 36.2% of revenue in the first quarter. We continue to make investments in labor and materials in support of R&D activities and incurred higher G&A expenses related to Singapore start-up and the legal actions to protect our intellectual property portfolio. Offsetting this increase were lower stock-based compensation as Q1 executive severance costs were partially offset by a new annual option grant.

Operating income for the second quarter was $23 million or 20.2% of revenue compared to $16.7 million and 16.3% in the first quarter. On a non-GAAP basis, operating income for the second quarter exceeded our targeted level at 25.8%, up from 23.8% in the first quarter. Volume leverage, factory execution and good products positioning all contributed to the improved performance. As anticipated, the tax rate was higher in Q2 to 34.6% principally due to planned start-up investments in Singapore and fewer stock exercises.

Net income for the second quarter was $18.6 million or $0.38 per fully diluted share versus $14.8 million or $0.30 per fully diluted share in Q1. Stock compensation charges were $6.4 million or $0.08 per fully diluted share for the second quarter compared to $7.7 million or $0.11 per fully diluted share in the first quarter which included $1.4 million related to a separation agreement.

Our Singapore expansion continues having signed a 30-year pre-paid land lease offer and moved into the design phase of building construction. In preparation for the next stage of anticipated growth, we are accelerating our plans by designing a combined clean room for the back-end assembly and test operations and front-end wafer fab in the first phase of construction. We'll bring the wafer fab online and install equipment as required potentially beginning 9 to 12 months after the back-end operations have begun at the end of 2008.

We are also planning additional clean room space for an advanced R&D lab for product and process development. We are enthusiastic about our expansion into Singapore and believe that it offers an ideal site for both back-end and a second wafer fab for FormFactor in the late 2009 early 2010 timeframe. As previously indicated, we expect to fund our expansion into Singapore with anticipated revenue growth within the ranges of our target P&L model on an annual basis.

In the second half we will begin funding part of our R&D investments out of Singapore. This early investment sharing between our Singapore and U.S. operations will position the Company to benefit sooner from lower tax rates in Singapore associated with our intellectual property assets. Now typically these start-up investments could be carried forward and credited against future Singapore taxable profits thus reducing the adverse impact on FormFactor's near-term tax rate. However, the favorable tax exempt status we have been granted in Singapore will likely make these loss carry-forwards unusable. As a result, our tax rate will increase to around 39% for the second half of 2007. A similar impact is expected to occur in 2008, as well. With the ramp of Singapore revenues in 2009, we will begin to benefit from the cost advantages of operating there most notably decreasing our corporate tax rate initially below 2006 levels, probably in the high 20% range. As volume continues to ramp in Singapore, the Company's tax rate will drop significantly. We see these steps as essential in planning for the next phase of FormFactor's growth and are confident that this project will give us even greater tax and other cost advantages over the long-term.

Now to the balance sheet, cash and marketable securities totaled $528 million in the second quarter, an increase of $20 million from the first quarter. Operating activities generated $34 million of cash in the quarter. DSO decreased to 41 days in the second quarter compared to 43 days in the first quarter as a result of improved shipment linearity. Net inventories increased by $5 million during the quarter to $32 million to accommodate increased customer demand for Q3.

During the second quarter $14 million was spent on capital expenditures for Livermore and regional facility upgrades and expansion. As I mentioned in the Q4 '06 earnings release, planned investments for Livermore and Singapore capacity expansion will likely result in capital spending higher than our 12% of sales target range. Given the current Singapore schedule including acceleration of the wafer fab clean room, we expect capital spending to be in the range of 13% to 14% of revenue for 2007 and potentially somewhat higher in 2008 depending upon revenue growth. Headcount remained flat for the quarter at 1,108 people. Direct labor and contractor reductions offset increases in sales, marketing and R&D headcount.

In summary, Q2 performance exceeded our expectations. Advanced probe card market growth and FormFactor's growth prospects remain solid and on track with our projections. Our technology position and depth has never been stronger although design resource limitations and Harmony ramp challenges have put us behind in our plans to expand into NAND Flash. As Igor mentioned, our overall technology and product differentiation remains unsurpassed and is progressing at a rapid pace with customers on several fronts as evidenced in recent press releases.

Now for our guidance for the third quarter of 2007. We expect Q3 to be another healthy growth quarter with revenue between $122 million and $127 million. We project non-GAAP operating income in the range of 25.5% to 26% and about 5 points of incremental stock comp expense. GAAP EPS should be in the range of $0.38 to $0.40 with non-GAAP EPS of $0.45 to $0.47.

With that we'll open the call for questions. Operator?

Thank you. (OPERATOR INSTRUCTIONS). Your first question comes from the line of Jim Covello with Goldman Sachs.

Great. Good afternoon, guys. Thanks so much. Congratulations on the good results. Question, first, can you order the drivers of growth for FormFactor for the second half of 2007 and for 2008? If you were starting with the drivers that are going to drive the most growth down to the least growth for the second half of '07 and '08?

So as I mentioned in the script, 70 nanometer transition is just beginning really and that is going to drive significant revenue growth in DRAM and it will basically -- you'll have proliferation of designs on 70 nanometer. You will have 1 gigabit devices ramping because that is going to be the sweet spot of satisfying Vista requirements and Microsoft Office 2007 requirements. You will have continuing growth of known good die revenue as it relates both to modal RAM and stack package execution as it relates to increasing adoption of known good die testing or some aspects of known good die testing in commodity DRAM.

You will have FormFactor penetrating new markets. As you know, we're working hard to gain market share both in Flash and in SOC. NOR market will continue to be healthy and with potentially some growth. So all of these elements will, we believe, will continue to drive our growth both in the second half. And going into 2008 Logic and SOC will become a bigger story.

So my follow-up would then be NAND Flash kicking in in 2007 is not required to hit the kind of targets that you're talking about for the rest of this year and the early part of next year?

We -- I would believe that, again, we will work very hard to continue working to NAND Flash, Harmony introduction and some of NAND Flash per my and Ron's remarks, some of which has to do with recent unprecedented design activity and we, again, are constrained in our design resources. Now that's something we're addressing, of course. We're not sitting still. But in those cases it does compel us to prioritize design resources to its customers and applications where FormFactor is mission critical where we cannot -- customers can't play without us and in some cases we're hear to not take certain NAND Flash business.

So all in all your comment is correct. We do not anticipate in the second half very significant NAND Flash growth but we are working hard to grow in the NAND Flash market over the long-term.

And we're seeing real strength, Jim, in our core markets, both our DRAM market and our existing Logic market, NOR Flash, all showing very strong through the year.

One more thing on NAND Flash is that NAND Flash is beginning -- there are new trends. Right? They are now specialty NAND Flash architectures that are evolving and beginning to compete for share in the cell phone market and trying to address an execution part of the stack packages; and for example, there is four NAND's out there and other new modifications of Flash that have higher read and write speeds. And those have higher probe card contained per bit in a way similar to how well known good die or mobile NAND business evolved where we have significantly higher probe card revenue per bit. So those markets are very interesting to us and we're focused on those as well.

Thanks so much. Congratulations, again.

Your next question is from the line of Gary Hsueh of CIBC World Markets. You may proceed.

Thanks for taking my question. My question here is based on margin performance and leverage that you can pull there. Obviously, it's a little disappointing that we've seen a lot of margin -- I'm sorry, multiple compression due to unclear, sort of, directionality here in terms of margins. I'm just wondering if you could kind of elucidate, what are some levers you could pull here not only in operating margin but specifically on gross margin that could significantly or perhaps improve moderately the drop-through on incremental revenue growth that, I think, we're all on board with?

I just wanted to start. Of course Ron will give you more color but all in all we're pretty comfortable with the financial model the Company's running with right now and that is as we've communicated on multiple occasions it's in the range of 53%, 55% gross margins. It's in the range of 25% operating margins with significant investments in R&D. And so that's, in general, the model how we would like to grow this Company. Remember this is a growth company.

Gary, we're in our model range and, as I mentioned, for a number of quarters now have been doing well. In fact, we had predicted that although Q1 was off our operating model range we would bring it back within the year and year-to-date we're already at 24.9% non-GAAP operating income, so sooner than we had originally profiled. And we certainly see strength continuing through the year.

As Igor mentioned, we are going to continue to operate in our model parameters and we are continuing to invest in growth for the future including the big move to Singapore which is being already reflected in our cost structure and will be through 2007 and 2008; but as I commented, we will on an annualized basis plan on staying within that target operating model construct and still fund all that expansion.

Alright. Thank you.

(OPERATOR INSTRUCTIONS) Your next question is from the line of Edward White with Lehman Brothers.

Thanks. I was wondering if you could talk about some of the things that you can do to get around some of the constraints in design resources and sort of the timeframe over which you think you can get to a point where you've got more design capacity.

This is Igor. As I mentioned, we are working them, Ed. Basically there are two ways. You hire more people and you automate your design tools and we're doing both. And the design demand is very, very high but we are both doing continuous improvement actions but we also are trying to think outside the box how to grow our design capacity over the short and long-term. So it's really adding people, adding them worldwide in different time zones and automating design tools.

So if you look at a timeframe over which you plan to sort of get around that, is it something that you can fix, say within the next few quarters or is it a gradual process? What kind of process is that and what sort of a timeframe you'd look at?

Well, the plan is to fix it in shorter timeframe than that. We do not want to be here within a few quarters and be in design constraint. Having said that, we are getting into new markets where there will be higher demand on our design capacity. So, in general, we will need to work both on continuous improvement action and step function increases in our design capacity in this Company. But we are working very hard pretty quickly (inaudible) capacity constraint.

Finally, related to that, will Singapore be able to help from a design constraint standpoint?

Yes. They will be a significant design resource. It's already been added in Singapore.

We've already added a few people, Ed. One point I'd just add to what Igor said is that when we're bringing on new products, not only do we have a ramp that goes on in our factory which we execute but we also have to bring on the design capabilities which, to start with, is more of a manual process. And, as Igor mentioned, you get it automated and you come up that ramp and then you have more efficiency. So when we're rolling out so many new products and technologies, we have that dynamic going on in the Company.

And one more regarding design, the new products -- when you introduce a new product platform, it's always been the case that at the introduction time it does draw significantly more per probe card design on the die capacity and then as the product becomes stable and introduced into the market it goes away. So there is a confluence of very high demand and confluence of having to introduce several new products at the same time that we're now working through. But I believe we're working through it well.

Great. Thank you.

Your next question is from the line of Harlan Sur of Morgan Stanley. You may proceed.

Thank you. Good afternoon and great job on the execution in the quarter. Igor, lots of new products, lots of innovation announced by the Company in the quarter. I know you formally introduced Takumi. We saw your CMOS image sensor card that Semicon West introduced a fine pitch Logic technology yesterday and a few others in the quarter. If you take all of this into account, I'm just wondering how much does this increase your dollar market opportunity over the next several years?

Well, this -- some of the new market segments that we're entering, for example, Logic SOC we're -- in our breakdown of different markets in the Company, say, Image Sensor would be a Logic SOC part of the business. And that is a very large opportunity. That's a very, very large market. So not only we are going to plan to gain market share and drive transition to this advanced technologies and Logic SOC but we also, by introducing this higher parallelism product such as image sensor and Takumi's of the first step in parametric you will see us doing pretty exciting things and in that side of the business where companies use our probe cards for (inaudible) learning and engineering and potentially in line measurements, some of those things amplify the markets. We really create markets. We transform markets. So, yes, all of this should bode well for our future growth.

Great. And then in the near-term in your core DRAM segment, your DRAM customers continue to focus on reducing test costs of the package although I think you might have touched them on this and they're really focusing on outputting ETT and UTT DRAM. And my sense is that this is putting more importance at the wafer level probe. I guess the question is, would you agree with this and is this helping to drive some of the demand trends in the second half of the year?

What's happening in DRAM, for example, when you do wafer level burning very early, not only do you reduce the package fall-out at final test, and package is 20% of DRAM cost, but also you can recapture some of the devices through wafer repair, right, as opposed to doing it at the end and not being able to do that.

So pulling more and more tests to the wafer, pulling more and more tests closer to front-end not only allows you to manage your supply chain downstream which means you manage packaging and you manage the package yield, but also it allows you to obtain very quickly, very valuable information during the times when people transition to new technologies at alarming frequency. So the speed of learning basically is separating people who potentially can do well in this market from people who won't. So we have a full suite of these products to allow companies to do entire validation of DRAM devices.

I'll give you one really exciting possibility. If you look at DRAM industry and you look at a huge portion of DRAM shaped as thin modules and if you open a DIM module under the (inaudible) you will find packaged DRAM devices that are poured on a little circuit board. And that's a DIM module. Well, if you validate your chips at the wafer level, you don't need those packages. You look at a company that spends $5 billion a year in DRAM costs and they spend $1 billion of packages so you could eliminate -- we're not talking here kind of small opportunities or borderline opportunities, we're talking in an industry that is struggling to make money, right? You're talking a huge, huge opportunity. So I believe that's the potential of moving more and more wafer tests -- more and more tests from what's to be the final package test to the wafer and it's beginning and it's taking time but when it's happening we will be a very strong beneficiary of that.

Thank you. That was very helpful.

Your next question is from the line of Mark Bachman with Pacific Crest Securities. Go ahead.

Thanks for taking my call. Ron, can you discuss this tax rate a little bit better? Can you give us guidance as if you didn't -- I'm assuming you're using the 39% tax rate for Q3? What would your guidance have been without that tax? I'm guessing about $0.04 higher? And just kind of a bigger question, was this a surprise that just came about? In other words, why the guidance now on this tax thing? How come we couldn't have heard about this sooner?

Mark, the tax rate if you look at the rate we had this quarter of about 34.6% would have been roughly what would have carried through the year. What we have done is actually accelerated our investments supporting the transfer or co-funding, if you will, of intellectual property assets out of Singapore. What that enables us to do is to get the benefits of the tax exempt status in Singapore on our intellectual property contributions sooner than we otherwise would have. So the way to think about it is that you pay a little more up front here in a few points of tax rate but as we ramp up volume in Singapore in 2009 and beyond, the tax rate will come down more substantially than it otherwise would have because we've got the intellectual property value in Singapore.

Now the other point I made in my comments is that, okay, up-front investments like funding R&D out of Singapore, normally you could put up on your balance sheet a deferred tax asset and then -- and that wouldn't affect your near-term tax rate and you take that against your profit when you start making profit in Singapore. The problem that you have is when your tax rate is zero, you don't have an asset that you can put up against taxable profit in the future. So that's a good problem to have but it does raise our tax rate in the short run.

And then the guidance without this tax would have been, what, roughly $0.04 higher?

In that range.

And then just my follow-up here. Ron, can you discuss the revenues by geography, here? There's a huge ramp in Japan and I'm just wondering, is this all from your largest customer or are you accounting for their [JV] in Taiwan through Japan or, by chance, is this a brand new customer?

In terms of the way we report our revenue in our reporting, we do report some of our U.S. -- or some foreign companies through their U.S. subs but that's mainly in Korea where we report on the U.S. numbers that we provide on the website, sales in U.S., that are actually to Korean companies through their U.S. subsidiaries. We don't have that same construct with Japan. So the Japan numbers are fairly clean in terms of the volume that we're actually shipping into Japan.

And is that from your largest customer?

Of course that involves our largest customer.

Does it involve more than your largest customer?

Yes. It can.

Thank you.

Your next question is from the line of Doug Reid with Thomas Weisel Partners. You may proceed.

Thanks for taking my questions and congratulations on the execution. Question about known good die and I'm trying to understand a little bit better the ramp for both wafer level burn and HFTAP. If you could help me understand just how lumpy that business appears to be as you look out two quarters and then if you could maybe characterize '08 to the extent possible in terms of recurring business there?

So known good die is driven by model RAM, NOR Flash and increasingly, as I mentioned before, you get DDR2 known good die business, mainly wafer level burning business. As I mentioned, Q4 is seasonally -- has been historically for us a quarter where consumer demand -- we are a quarter ahead of the end markets kind of or something like that. So Q4 has been a softer quarter for us in model RAM and consumer -- in the consumer market bound devices that are tested without a probe card. But longer term outlook to '08 and beyond for KGD is very bright.

Just as a follow-up to that, it's my understanding that as you look back to '06 that seasonal decline in Q4 in part related to the high concentration of revenue at a single customer. How is revenue branched out or what is the concentration, rather? What are the top two customers in percent of total?

Well, we are now shipping known good die to multiple customers so my remarks as to Q4 potential of being soft, again, for a consumer market bound device tested with our probe card is -- has to do with the industry and seasonality trend, not single customer remark. We are shipping known good die probe cards for high frequency test and wafer level burning to model RAM. We're shipping high frequency test probe cards to NOR Flash customers and it's a pretty large business for us and we're shaping wafer level burning probe cards to commodity DRAM customers and we're talking multiple customers. So this is now spread over many customers.

That's helpful. Thank you.

Your next question is from the line of Tim Arcuri of Citigroup. Please proceed.

Hi guys. Igor, you've been talking about the industry growing about 25% this year and I think you before said that you would grow at least 25% this year. If I just take the midpoint of your revenue guidance and if I assume that you're going to grow about 25% this year, that implies that revenue in Q4 is flat to down. So with respect to what you think you'll grow this year relative to the market, what would your comment be there?

This year we plan to grow 25% or over 25%.

So with respect to the fourth quarter, it's -- at this point it looks kind of down, maybe flat-ish?

We -- Tim, as you know, I'm not trying to avoid your question but we only guide next quarter.

Of course. Okay. Thanks.

Your next question is from the line of Mark FitzGerald with Banc of America Securities.

Thanks. Igor, I understand the kind of design issues on the NAND side of the business but on the manufacturing side, what are the challenges you've got to solve here to get this product up and ramping and accepted out there?

Yes. As we have been pretty public with the fact that we are behind Harmony in Harmony introduction as compared to say a year ago where we sat. We wanted to be, of course, farther ahead than we are today. We have made very significant progress in the last industry. I mean, we had a mighty amount of resources. By mighty I mean mighty by the standards of probe card industry. We had significant focus and significant very brilliant resources on this and we're making progress. And the main issue is making this Harmony platform working for multiple testers. And it has to do with test cell integration and it has to do with fitting into a Legacy interface design that was not designed for such a product. And that's what we're working through.

We believe we will work through this. We believe that we will be very, very well positioned for new tester platforms that will enter the market. For example, if you want to do two touchdowns or one touchdown in DRAM, it will be a (inaudible) platform. So -- and we just believe that what we're learning right now has to do with integration into Legacy tester platforms.

And just a quick follow on these, any issue in some of these newer markets like the Logic SOC or any of the other markets you're addressing in terms of having to go back and fit it into a Legacy that you face today in Flash and --?

Those are very different probe cards. In SOC we're introducing probe cards that can do 32 or 64 devices in (inaudible) testing which is completely revolutionary for SOC and we'll do this complete marvels to the wafer test economics in SOC. But this is only 1/4 or 1/8 of the (inaudible) in the testing areas, right, that we do in DRAM. So, no, our challenges with Harmony are not -- should not be indicative of what we might experience in SOC platforms. Having said that, it will require now that you have many more tester platforms so you need engineering and you need design resources to get those things on the SOC side done and that's what we're focused on. We formed separate business units. We have separate focus teams now both from business standpoint, engineering execution standpoint, focusing on these different market segments and customers and we are making progress in SOC.

Thank you.

Your next question is from the line of Patrick Ho of Stifel Nicolaus. You may proceed.

Thanks a lot and congratulations on a nice quarter. I think following Mark's question about the NAND Flash Harmony product, how much would you say is a combination of both your own, like for the die capacity shortages that you're experiencing versus the customer interest? It sounds like -- is manufacturing the biggest issue that you're facing or is it the other factors as well?

We -- there are -- we're working through Harmony, new Harmony platform introduction. Those are issues -- again, we're making progress on those. It's happening slower but we're making progress and we made good progress during Q2. In one case the product is running in high volume production and it was benchmarked against alternatives and it is very well in several areas. It is a better product.

We have overall design constraints right now due to very, very high design demand in general. And as you can see we're growing our business at a healthy chunk and that puts significant pressure on design resources.

Under our situation, when we're faced with a choice of where to prioritize our design resources, believe me we will always prioritize it for cases when customers absolutely depend on us to run their fab as opposed to cases when we are working hard to gain market share in their market. So that was the nature of my remarks and as we have more design capacity, we believe we will be gaining more market share in NAND with time.

Great. A final question on the KGD market, where do you -- it's obviously grown pretty fast over the year so far. Where do you see it by the end of 2008? How much of a percentage of revenues do you think you'll get from the KGD market?

Well we, again, we see this as a significant growing market for FormFactor so you should see -- if I had to guess -- between 30% and 40%. But this is not a number that we're tracking on a regular basis. I mean, looking at the year, year and one-half ahead. It's not something that we're looking at all the time. But if I had to guess -- 30% to 40% of revenue.

And your final question comes from the line of Colin McArdle with Needham & Company.

Hi guys. I believe on the last quarterly call you stated that the manufacturing capability in Livermore could sustain $125 million in quarterly revenue and based on the high end of your current guidance, obviously you've increased that and I wonder what that number was currently and where you hope to see it by year-end?

Hey, Colin, this is Ron. We talked about the current capacity last quarter of our fab at about $130 million of capacity. We are able to increment that capacity step-wise by adding machines in bottleneck areas step-wise as we grow here through 2007 and 2008 prior to bringing up Singapore. So we've got plenty of room to expand into next year but the current capacity, as I mentioned last quarter, is about $130 million. We're a little bit above that now and we'll keep scaling that as we need.

Actually the scaling of capacity is one of the areas -- enormous amount of work and planning has been put into that and we believe we can scale capacity successfully over the next - manufacturing capacity - over the next many years ahead of market demand. We will -- the way it's designed, actually, a lot of brilliant work went into that and for us -- the lead times increased capacity; also something that we understand much better and we can react to market conditions. So we are very confident we can grow capacity with demand or ahead of demand.

So is it fair to say as a good quarter unfolds you could add $5 million in capacity just by tweaking Livermore fab?

Our manufacturing guy will probably choke me but the answer probably is yes.

Tweaking. The emphasis is on tweaking.

Thanks, guys.

And at this time I would like to turn the call back to Mr. Ron Foster for his final remarks.

I want to thank you all for joining us today. We certainly enjoyed it and we look forward to seeing you at our future conferences and on our next earnings call. Thank you very much.

Ladies and gentlemen, thank you for your participation in today's conference. This does conclude the presentation and you may now disconnect. Have a wonderful day.