FormFactor Inc (FORM) 2006 Q4 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the FormFactor fourth-quarter 2006 earnings conference call. My name is Oneka, and I will be your operator for today. (OPERATOR INSTRUCTIONS). As a reminder, this conference is being recorded for replay purposes.

At this time, I will now like to turn the call over to Ms. Anne Leschin. Please proceed.

Good afternoon, and thank you for joining FormFactor's fourth-quarter 2006 earnings conference call. With me on today's call are Igor Khandros, CEO, and Ron Foster, Chief Financial Officer. I will provide a summary of our fourth-quarter performance, review our market segments, and provide an update on outlook and long-term strategy. Ron will then take us through the financials, operational details, and provide guidance.

I would like to take a moment to mention that during the first quarter of 2007, the Company will be presenting at a number of conferences, including Thomas Weisel Partners Technology Conference on February 5 in San Francisco at 2:40 PM at the Fairmont Hotel; the CIBC World Markets Semiconductor Summit on February 22 in Vail, Colorado; the Goldman Sachs Technology Investment Symposium on February 28 in Las Vegas, Nevada; and the Morgan Stanley Technology Conference on March 7, also in San Francisco. 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 call, we will make projections within the meaning of the federal securities laws, including statements regarding FormFactor's growth and financial performance as well as our strategic and operational plans. These 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 extent to which chip manufacturers modify announced capital equipment expenditures and device roadmaps; and the Company's ability to efficiently execute on its new product introductions; customer selection of these new products; and capacity expansion plans.

Please refer to the Company's recent filings on Form 10-K and 10-Q for more detailed discussions of 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 and bookings by market and geography and a GAAP to non-GAAP reconciling income statement are available on our website, as is a schedule reconciling our GAAP and certain non-GAAP financial guidance with respect to FAS 123R stock compensation expensing.

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

Capped by record revenue of $98.7 million in fourth quarter, 2006 was a milestone year for FormFactor. A year ago, we set an ambitious list of goals for the Company -- to successfully complete the transition and ramp production in the new factory; to outgrow the market while elevating customer satisfaction metrics to historically record levels; and to introduce an unprecedented number of new products -- all of which we successfully accomplished during the year, and will fuel our growth in the next several years.

In 2006, FormFactor grew 55% to reach $369 million in revenue, once again outpacing the advanced probe card market, which we estimate grew 40%. Our 2006 growth rate, as well as our three-year CAGR of 55%, far outperformed the semiconductor and semiconductor capital equipment sectors. Operationally, our factory performance vastly improved, as we increased on-time deliveries and reduced lead-times, culminating in better customer satisfaction and as a result, setting the stage for expanded business relationships with existing and new customers.

We solidified the globalization strategy of FormFactor with our previously-announced plans to expand back-end manufacturing in Singapore and the introduction of the [Kumi] product line, which was developed and is managed entirely by our Japanese team. We introduced a total of six new products, including two new platforms, representing an unprecedented level of innovation for the Company and our industry.

These new product introductions did not come without their challenges. Most significantly, it took us longer than anticipated to reach volume shipments of our Harmony NAND Flash product. Although currently, our product ramp is tracking with the migration plans of the early adopting customers.

Following the demand surge in the first half of 2006, the market for advanced probe cards flattened in the second half of the year and in the fourth quarter specifically with the seasonality of the consumer market and the slower transition to 70 nanometer. FormFactor's growth in the fourth quarter was moderate and driven largely by record SoC and NOR business, as well as growth in DDR2, 80 nanometer, and 300 millimeter.

Qualifications of our Harmony OneTouch NAND Flash product continued in the quarter with three customers, and we began volume shipments to one early adopter. DRAM remained a significant driver of FormFactor's growth in 2006 due to the normal seasonal slowdown in the consumer segment, model ramp, KGD and customers delaying their transition to advanced nodes, such as 70 nanometer, and new products, such as 1 gigabit device for Vista. Our DRAM business slowed slightly in the quarter.

With the rollout of Vista, we are now witnessing an accelerated transition to 70 nanometer, supported by recent customer announcements on improved yields at 70 nanometer, and resulting in a noticeable increase in our design win and order activity. We currently have customers successfully qualified in our 70 nanometer products, and anticipate a significant and steady increase in order activity in the first half of 2007.

Significantly, customer forecasts currently point to shipments increasing throughout the year. The industry view is that big crossover to a combined 80/70 nanometer node will be in the third quarter, and we will typically see our probe card business ramp one/two quarters ahead of this crossover. Along with 80 nanometer growth, we see the DRAM 70 nanometer transition as a major probe card tooling event in 2007 and 2008. Tooling cycles typically last two years, and we expect the total 70 nanometer tooling event revenue to grow 50% over the prior tooling cycle, as in past transitions.

Earlier this week, we introduced our Harmony XP product for DRAM, with orders for this product from two customers. This leading-edge product was designed to address the technical challenges of high-density mobile, commodity and graphics DRAM devices, and to leverage the resources of state-of-the-art testers, which can test record 384 devices concurrently and test a 300-millimeter wafer in as few as two touchdowns. We believe Harmony is the platform that will meet the advanced test requirements of full 300 millimeter wafer testing, and sub-70 nanometer designs at the lowest total cost of ownership, and will extend our lead as the premier supplier of advanced probe cards to the DRAM industry.

2006 has been a pivotal year for mobile RAM and KGD products, as mobile RAM revenues doubled and KGD revenues tripled. This is despite the slower fourth quarter, where we experienced the expected seasonal demand decline in this consumer market-related product.

Mobile RAM will follow standard PC DRAM in migration to 80 and 70 nanometer nodes, with the big growth projected at 90%. Consequently, we expect probe cards for consumer memory applications to ramp to a peak at mid-year, along with the planned migration of DRAM to 70 nanometer. The increasing requirements for known good die testing in multiple market segments -- mobile RAM, PSRAM and NOR -- will drive our business and the segment. As a result, we see more device manufacturers entering the KGD market in 2007 and we plan to expand our customer base.

NOR Flash showed significant growth in the quarter, primarily due to the strong production expansion of one top NOR customer. Our focus in delivering our Harmony OneTouch NAND product to the market progressed with three key early adopter customers. As we announced today, we received a large order from Hynix for our Harmony OneTouch NAND product, which also positions us well for their future Flash business. We also have two other customers currently qualifying our one-touchdown Harmony OneTouch card, which we expect to lead to volume orders.

Designed specifically for the cost-sensitive or productivity-focused test environment of the NAND market, our Harmony OneTouch NAND product is exhibiting key differentiated performance results and early qualifications in 200 and 300-millimeter applications by providing our customers with higher uptime, faster setup, shorter lead-times, lower cost of maintenance, and as always high-quality electrical performance. To date, the NAND market has had less complex test requirements than DRAM market, as NAND speeds increase and electrical characteristics and probing precision become more important over time. Harmony OneTouch is well-positioned to additionally differentiate and become the high-volume product of choice at NAND market leaders.

We are watching carefully the recent weakening in NAND Flash market, which is causing several customers to direct their new memory capacity to DRAM. We see this as a positive for FormFactor, as it moves the customers to the sweet spot of FormFactor's established product offerings and affords us extra time for our OneTouch NAND RAM.

In addition to the NAND growth opportunities, we will also capitalize on the increasing KGD test requirements for NOR Flash, leveraging our highly-differentiated, high-frequency test at probe products. All in all, we expect to continue our steady growth in Flash as our new products continue to redefine the market in 2007.

FormFactor's SoC business reached record levels in Q4 and in 2006, driven primarily by key microprocessor customer in our flip-chip market. We also had a key customer qualification in the gaming segment, and look for business RAM from this customer by mid-2007.

The development of our fine-pitch SoC product continued in the quarter through our work with one early adopting customer in the high-parallelism wire-bonded fine-pitch solution. In 2006, we developed an entirely new spring and product architecture for this market. In 2007, we plan to commercialize this product as wire-bonded SoC customers using predominantly needle cards convert to our higher parallelism probe cards.

In 2007, we see an inflection point in SoC market as customers are starting to request higher levels of parallelism in products ranging from microcontrollers, ASICs and DSPs to chips for cell phone, automobiles, game consoles, and in image sensors for camera phones. Today, generally, only one or two SoC devices are tested in parallel versus 256 or more in DRAM.

Until recently, SoC testing did not require high parallelism probing capability. But the integration of embedded Flash memory into SoC, for example, is driving up test dies. New solutions are needed to close the functional gap of legacy testers to address increased volumes and higher test dies, just as memory testing evolved to higher levels of parallelism, so too will SoC testing.

Moving to probe card technology that enabled the testing of four or eight devices in parallel will dramatically reduce manufacturers' capital expenditures. We believe that the system level approach that involves closer collaboration between [80] suppliers, device engineers and the provider of probing solutions will drive the transition to higher parallelism testing in SoC in the next two years, something that has been a key part of FormFactor's strategy.

We ended 2007 as the leader in a growing market. FormFactor estimates that the 2005 to 2010 CAGR for the overall market, including probe cards and solutions, will be 30%. In 2007, we expect the advanced probe card market to increase roughly 25%, following the estimated 40% increase in 2006.

Certain market segments will outperform the growth rate next year, such as DRAM and SoC. While others, such as NAND, will grow at the slower rate. With this market picture, FormFactor continues to believe that it is capable of outpacing the market in 2007. While the probe card market will provide growth dollars for multiple players, we believe we will strengthen our position as the primary mission-critical supplier in the advanced probe card market, while some competitors may benefit from growth primarily as second sources.

Now, I would like to turn to our 2007 strategy and the major market drivers we see by segment. In summary, 2007 growth will be achieved by strengthening our DRAM market leadership, by adding new flip-chip SoC customers, and by gaining market share in fine-pitch wire-bonded SoC and NAND markets with new products.

Our outlook for the DRAM market is very healthy as we see 2007 as a year for PC server DRAM, though consumer DRAM will grow as well. The key drivers are strong bid growth at 65% for DRAM, the transition to 70 nanometer which has begun, the migration to 1 gigabit devices with Vista's release, strong model device growth leading to 90% plus bid growth in mobile RAM and increased adoption of known good die testing.

As we have maintained over three years, known good die testing will be one of our key markets as an increasing percent of testers move into the wafer. We believe it will grow and amplify our current markets in 2007 and beyond, driven by stack package requirements for cell phones, with mobile RAM, PSRAM and NOR Flash, and with the inevitable long-term trend of packaged to wafer test transition in commodity DRAMs.

In Flash, we expect moderate market growth in 2007 and we have the opportunity to grow faster than the market. The primary drivers are health in the NAND industry bid growth, increased NAND market share based on the penetration of our Harmony OneTouch product, the NAND technology transition from 70 to 50 nanometers, and continued command of the NOR market by FormFactor.

In SoC, FormFactor plans to usher in the era of higher parallelism wire-bonded fine-pitch probing in 2007 by enabling the conversion to advanced probe cards in this market. We will expand our strategic customer engagements during the year.

Across all markets, FormFactor's products offer superior performance in their productivity, yield and parallelism. The breadth and depth of our technology and product offerings combined with the strategic advantage of our factory and our worldwide applications and customer service will enable our memory and SoC customers to keep pace with their technology and course roadmaps and FormFactor's continued success in existing and new markets.

To summarize, 2006 was another remarkable year for FormFactor with 55% growth, well above our estimate of the industry's growth. We are pleased with the Company's fourth-quarter performance as we achieved record revenues, expanded our manufacturing and order fulfillment capability, and introduced a number of revolutionary new products and architectures, setting the right foundation for a successful 2007. We believe that FormFactor is well-positioned to capitalize on the many market opportunities in front of us, and can outgrow the market in 2007.

Finally, I would like to thank the entire FormFactor team for their hard work and efforts to make this another successful quarter and successful year for FormFactor. I will now turn the call over to our CFO, Ron Foster, who will elaborate on operating results, financial performance for the fourth quarter, and provide guidance for the first quarter.

As Igor mentioned, 2006 was an exceptional year for us from both an operating and financial performance perspective. We grew our revenue 55% in the year to $369 million. DRAM revenue continued to drive our growth, increasing 49% in the year to $272 million, accounting for 74% of our revenues in 2006. We expanded in the Flash market, increasing revenue 84% to $58 million, or 16% of total revenue for the year, and increasing our base of customers. Logic revenue grew 69% to $39 million, representing 10% of total revenue in 2006.

We made significant gains in factory productivity, yields, and on-time delivery. These factory improvements enabled us to increase capacity and improve gross margin from 45.2% in 2005 to 52.8% on a comparable non-GAAP basis, excluding FAS 123R option expense.

Factory capacity has increased from $80 million per quarter as we exited 2005 to $125 million per quarter today, greater than 50% increase in one year. This capacity increase has been achieved through improving yields, throughput productivity, cycle time reductions, and capital investments.

One manifestation of faster factory cycle time is the rate at which orders can now be booked and turned in the same quarter. The turns ratio has increased from the low 40% range at the beginning of the year to about 60% in the fourth quarter. As we continue to improve productivity, we expect our turns business to comprise a larger portion of revenues, making bookings correspondingly less of an indicator of future revenues. Consequently, comments about bookings going forward will be focused on instances where it provides useful market insight.

We also made progress in planning for strategic capacity expansion in Livermore and Singapore. Once implemented, these expansions will significantly increase capacity, lower our cost structure, and position the Company to maintain healthy profitability. The plan is to ramp capacity in Singapore in the second half of 2008. Capacity in Livermore can be increased to meet demand until Singapore is in operation.

We are in the process of identifying real estate in Singapore that will meet our needs for at least the next five years. We have finalized an agreement for tax incentives with the Singapore government that provide significant tax rate reduction for work performed in Singapore. We are also completing arrangements for the government for support with workforce training and R&D incentives. The improvement in gross margin enabled increased investment in research and development programs and infrastructure to support further growth.

Operating margin increased from 14.9% in 2005 to 19.7% in 2006 or a 25.1% when adjusted for option expenses not included in the 2005 numbers. Operating profit dollars on a comparable non-GAAP basis grew 162% year-over-year on 55% revenue growth. We achieved our target annual operating profit model of 25% for the first year in our history. Going forward, we expect to maintain this performance on an annual basis, and support 30% compound annual revenue growth.

Earnings per fully diluted share was $1.29 in 2006 or $1.58 excluding FAS 123R option expense, which compares to $0.73 in 2005. 2006 ended with $495 million in cash and investments, an increase of $281 million compared to year-end 2005, due largely to our $182 million follow-on offering in Q1. Stronger earnings and cash management added $103 million from operating activities, and free cash flow was $65 million.

I will now provide specific comments on our Q4 results. Revenues once again hit record levels this quarter, climbing for the eighth consecutive quarter to reach $98.7 million, 2% over the third quarter and up 37% versus Q4 of 2005. Revenues grew significantly in Taiwan, where our business with key foundries continues to expand, and in Europe with a tooling event with a major memory customer.

Revenues in Japan decreased from the third quarter's record levels, due largely to the completion of a customer's tooling cycle for 90 nanometer mobile RAM. Revenues declined slightly in both North America and Korea as a result of lower DRAM business, partially offset by NOR Flash growth.

DRAM revenues, which accounted for 69% of revenues in the fourth quarter, declined 2% sequentially to $68.4 million, due to a sharp decrease in mobile RAM business, in line with expected seasonal trends. DDR2 revenues were strong, increasing by 32% from the third quarter, as DRAM manufacturers continued to benefit from high DRAM prices and the broadening of DDR2 to 128/256 megabyte and 1 gigabit designs.

Certain customers also ramped 80 nanometer DRAM, increasing our revenues and advanced nodes in the quarter. Approximately 90% of DRAM business was generated by 90 nanometer or below, compared to 86% in the previous quarter.

Flash revenues increased 1% sequentially to $16.9 million, representing 17% of revenues in the fourth quarter. Flash grew 109% over Q4 2005.

NOR revenues achieved record levels in the quarter, primarily from one customer's tooling cycle. NAND revenues decreased sequentially as we migrated from [NF150-based] designs, and began shipping initial volumes of our new Harmony OneTouch product.

Logic revenue was $13.3 million, representing 14% of revenues, an increase of 37% over the third quarter and 47% from the fourth quarter of 2005 due to the growth of a key customer in our microprocessor and chipset business and growth in our customer base.

Known good die revenue, consisting of our wafer-level burn-in and high frequency tested probe HFTAP products, tripled in 2006 to $40 million as consumer applications became the early adopters of this accelerating trend. The fourth quarter was a seasonally-weak consumer market for wafer level burn-in and HFTAP products, as it was with mobile RAM, with a 50% decline to $6.7 million from the lofty Q3 level. The breakdown of revenues by market and geography is available on our website.

Fourth-quarter bookings were $91.3 million, an increase of 1% over the third quarter, and 11% over the same period last year. Our schedule of bookings by market segment is available also on our website.

Now, I will discuss our GAAP results and some key non-GAAP results to supplement understanding of our financials. A schedule that provides GAAP to non-GAAP reconciliation is available on the investor portion of our website.

GAAP gross margin for the quarter was 51.8%, slightly below the 51.9% for the third quarter. Fourth-quarter gross margin included a charge of $1.3 million, or 1.4% of revenues, for stock-based compensation attributable to FAS 123R. This compares to a $1.2 million charge, or 1.3% of revenues, in the third quarter.

Non-GAAP gross margin of 53.2% was flat with last quarter, and in line with our target range of 53 to 55% for the third consecutive quarter. Labor and material costs improved in the quarter as a percent of revenue, and we continued to add capacity to the factory.

GAAP operating expenses for the quarter were $31.7 million, or 32.1% of revenue, compared to $31.3 million, or 32.4% of revenue in the third quarter. FAS 123R option expense added $4.6 million, or 4.7% in the quarter, compared to $4.5 million, or 4.7% in the third quarter. R&D expenses increased to $13.2 million, or 13.4% of revenues, versus $12 million, or 12.4% of revenues last quarter, as we continued to invest in the development of our Harmony architecture, fine-pitch logic and other new products and technologies.

Non-GAAP R&D expenses were $11.7 million, or 11.9% of revenues. SG&A expenses decreased $815,000 to $18.5 million, or 18.8% of revenues for the quarter, compared to 20% of revenues for the third quarter. Litigation and compliance expenses for the quarter were $1.7 million, compared to $2.1 million in the third quarter.

Operating income for the fourth quarter was $19.4 million, or 19.7% of revenue, compared to $19 million and 19.6% in the third quarter. FAS 123R charges decreased operating income in the quarter by $5.9 million, or 6%, compared to $5.7 million or 5.9% in the third quarter. On a non-GAAP basis, operating income for the fourth quarter was 25.7%, achieving our annual target model range for the quarter as well as for the year.

Interest and other income for the fourth quarter increased to $5.1 million from $4.5 million last quarter. The increase was attributable to higher cash balances and higher interest rates. The yield on our cash investments in the quarter was 4.3%, with a mix of tax-exempt and taxable investments.

The effective tax rate for the quarter was 23% compared to 32.7% in the third quarter. In December, Congress passed federal tax legislation leading to the retroactive reinstatement of the federal R&D tax credit for 2006 and prospectively for 2007. Excluding the federal tax legislation, the effective tax rate for the quarter would have been 35%. The tax rate for 2007 is projected to be 33%, with the benefit of the R&D tax credit.

Net income for the fourth quarter was a record at $18.9 million and $0.39 per fully diluted share on a GAAP basis compared to $15.8 million and $0.33 per fully diluted share for Q3, due partially to the tax rate adjustment I just mentioned. FAS 123R resulted in a net charge to earnings in the quarter of $4.6 million, or $0.09 per fully diluted share, compared to $3.9 million dollars, or $0.08 per fully diluted share in the third quarter. The $0.01 increase in FAS 123 expense quarter to quarter was entirely due to the lower tax rate in Q4. With the tax rate normalized to 33%, Q4 GAAP EPS would have been $0.34.

Cash and marketable securities totaled $495 million in the fourth quarter, an increase of $35 million from the third quarter. Operating activities generated $44.3 million of cash in the quarter, compared to $17.5 million last quarter. Cash flow from operations increased due to working capital improvements, notably receivables and accounts payable.

During the fourth quarter, $11 million was spent on capital expenditures, compared to $8.3 million in the third quarter. The majority of which was associated with expanding facility capacity and new product technology. For the year, capital expenditures were 10.3% of revenue.

In 2007, capital expenditures could be higher than 12% annual target range depending on revenue level for the year, as we expand investments in Livermore and Singapore. We will consider purchasing rather than leasing some of our facilities, where it improves overall financing economics.

DSOs declined to 37 days in the fourth quarter compared to 42 days in the third quarter. The accounts receivable balance increased 474,000 from the end of the third quarter.

Net inventories decreased by $1.5 million during the quarter to $24.8 million. Inventory turns were $8.3 in Q4 compared to $8.4 in Q3.

Headcount decreased to 1047 at year-end, down from 1052 in the third quarter, as product development, customer support and other infrastructure headcount investments increased, while manufacturing headcount declined in the third quarter. Now, for our guidance for the first quarter of 2007.

We expect revenues to be $98 million to $102 million. We project operating income in the range of 14 to 16%, including about 7.5 points of incremental FAS 123R stock comp expense. The non-GAAP operating income will be in the range of 21.5 to 23.5%.

This guidance includes about $1.8 million for the separation agreement with our former President, $1.4 million of which relates to FAS 123 option expense. Operating income will be additionally impacted by $1.4 million in higher employee benefit costs that are typically recorded at the beginning of each calendar year, and then decline as a percentage of payroll in subsequent quarters.

Investments in plant capacity, new product technologies, and global infrastructure are also accelerating as we prepare for revenue and customer installation ramps throughout 2007. These items are impacting the operating profit for the first quarter. However, we expect to stay on our annual target model of 25% non-GAAP operating margin for the fiscal year 2007 as we did in 2006.

We target GAAP earnings per share of $0.26 to $0.30 per fully diluted share. This includes a $0.10 impact from FAS 123R expenses related to stock-based comp in the GAAP results. So non-GAAP EPS is projected to be $0.36 to $0.40. A schedule reconciling GAAP to non-GAAP guidance is available on the website.

Now, let us open the call for questions. Operator?

(OPERATOR INSTRUCTIONS). Jim Covello, Goldman Sachs.

One question on the financials and then one question for Igor on the NAND Flash market. Ron, can you give us -- I know you don't give explicit gross margin guidance for the out quarter. But can you give us some color on how much of these charges are going to show up in gross margin versus below the line?

Yes, Jim. As you observe, we don't try to break down gross margin with that degree of precision on the outlook quarter. But what I can do is give you sort of a characterization of what is impacting the margin.

If you look at it on a non-GAAP operating margin basis, the fringe and separation effects are impacting about 2 points on our operating income. The fringe effects are spread pretty much along the lines of the payroll structure. So it hits both areas, and that is a seasonal effect, as I described.

Also, there is a couple of points related to investments, as I mentioned in my comments. We are expanding in the applications area in design resources, which of course hit our operating expense categories. And those are to build out our structures to support customers and to improve our product offerings and helping them with their test time reduction.

We also are in the process of refining our factory layout as a first step in moving towards our Singapore back-end operation. And we will have a dedicated wafer fab operation in Livermore and a dedicated back-end operation in Livermore. And so those steps are starting in the factory. So those do have gross margin effects.

Then, we are also investing in new product activities in our factory, including the NAND platform and also the fine-pitch logic product, which will be rolled into our factory. So those will have effects on our cost structure there. But in general, I think we will be in good shape to maintain our gross margins going forward and in good shape in the first quarter as we trend through the year.

Then if I could ask Igor, on the NAND market, you have talked a lot about how -- Form's competitive advantage in NAND probe cards is going to kick in as NAND speeds increase. Can you talk a little bit about what drives the NAND speed increase, or how you see that progression developing so that you guys can have the potential to have the same kind of share in NAND that you have in DRAM today?

Yes, so on NAND, the speed increase is a longer-term trend. And there will be designs out there, where people put a DDR-like interface on NAND (indiscernible) applications and all the other things and then in general, will increase in speed. What will also happen as memory makers shrink NAND and those chips become smaller, you will have significant requirements for probing precision that is not there.

So I think you will see that starting in 2007, the way we look at it, the market for one-touchdown NAND should really shape up and solidify towards the middle of this year. We previously had thought it would be a quarter sooner. But with the recent direction of additional capacity by memory makers from NAND to DRAM, it looks like that market will really take on its important shape in the middle of the year.

As I said in my remarks, it is affording us some more time to make sure that we prove on all of those test floors where we are right now the advantages of our product. And we -- as again I mentioned in my remarks, we feel pretty good about it. We now are on high-volume production with one customer. We have very positive qualification results at other early adopters. So we feel that it will shape well for us.

Terrific. If I could just ask one quick other one. You mentioned a couple times, both in your prepared remarks and just there, the transition from NAND to DRAM capacity by some of the chip makers. I think we all had heard a little bit of that from some of your customers. And most of that was capacity that was planned to come up as NAND kind of got switched before it went into production, got switched to DRAM.

Are you kind of suggesting that there might be some more of that going on than what we have heard about publicly?

I am not suggesting anything that you can't find out from memory makers. I think -- it is a lot more difficult than previously thought, as we understand, to just shift capacity back and forth between NAND and DRAM. That really has not transpired.

So what has happened in the new capacity, with the kind of pricing and profitability picture that exists in the industry, the new capacity that you just mentioned is getting directed to DRAM. And there are evidences in several new fabs for that. Again, we're watching this very carefully, and it is kind of new news. It is just a couple of months old, and it is continuing. So we will see how this shapes up.

But to add to that, we are not committing hara-kiri here. Our customer is shifting NAND Flash capacity to DRAM. DRAM is where we have leading products that are well-established, and we feel actually pretty positive about it.

(OPERATOR INSTRUCTIONS). Mark Bachman, Pacific Crest.

Two quick questions for you. One is, if you can address in detail the demand for DDR2 designs with respect to both competition and what effect the recent price declines have had in this segment, or in terms of how you view orders going forward?

The second thing is, can you please address also in detail whether the issues that you incurred before Q4 with regard to planarity on your NAND Flash, has that been solved in its entirety for all OEM-relevant test platforms?

Okay, so the first question is about DDR2 and competition. Really, the story in DRAM this year is not the architecture. The architecture story, the DDR3 will probably start in earnest in 2008. The story this year are technology transitions to 80 and 70 nanometers.

70 nanometers was a big, a very significant tooling cycle for FormFactor, as again I mentioned earlier, we see -- we estimate it will be about 50% larger than our 90 nanometer tooling cycle. And the 80 nanometer kind of half-node is actually incremental business, because customers -- most customers who will have 80 nanometer, they will also have 70 nanometer. So that is really a big story.

As far as our competitive position in DRAM, we are -- we feel that we actually strengthened our position in '06, and we feel confident in '07. Customers always look for second sources. They always do. The customers always would like -- now, it doesn't mean that all customers invoke second sources at all times. But memory makers, as any other company that runs manufacturing, would like to always have a second source.

The second sources over the history have changed. It used to be [Gem] was vying for being such a company. DRAM today, probably the largest say of second source business belongs to Phicom in Korea. And that may change in the future.

And it looks like MJC is making a dent. But we again are very confident about our competitive situation in DRAM, which is also significantly, I believe, strengthened by our recent announcement -- I believe it was yesterday -- of Harmony XP, which really widens, I believe, the gap between FormFactor high-volume manufacturing capabilities and that of the fragmented competitive field.

So Igor --

The second question was on NAND Flash. And as we reported, I believe, in Q3 earnings call that we did in October that we did experience issues with one tester platform on integrating our OneTouch NAND solution with the tester interface. That problem has been solved. And we actually right now in that particular case, we are beginning to test production wafers. So that problem does not exist.

And we also integrated on other platforms now. So this is not a major issue facing us now. It is not even an issue facing us right now. It's basically work right now, and making sure that customers run our product in high-volume manufacture.

And then, just Igor, on the DDR2 side, are the recent price declines pushing a faster adoption of 70 nanometers for you?

Yes, sorry; I missed that part of your question. When we were very carefully looking for the onset of 70 nanometer transition, when we tried to forecast it, then this issue of pricing was one of the several issues, where we were watching very carefully.

But as I said in my script, now we don't need to forecast it. I mean, we have designs in-house. As Ron in his remarks mentioned, we are actually busy investing external sources in design to cope with design demand. So as evidenced by orders and design activity, the 70 nanometer transition is beginning in a big way.

Harlan Sur, Morgan Stanley.

Great job on the fourth-quarter execution guys. Igor, things have certainly changed for it, as you said from a quarter ago, when it was looking like technology transitions in DRAM were at a standstill. So things have certainly accelerated more recently. Clearly, we have heard from a number of your customers about accelerated node shrinks.

And so as we look at the first quarter, I am just -- and I think you might have touched upon this -- but has the business seen an acceleration of order activity quarter to date here in Q1 because of the pickup in technology transitions?

I think that the best answer to that question is the revenue guidance that Ron gave. And clearly, we see a strong outlook from here moving forward and a strong outlook for 2007.

(multiple speakers)

-- about it.

There is one thing I might just add to that is that I mentioned about the design investments. And we are making significant investments in design activity, which is the front end of our ordering and manufacturing process, if you will, because we are seeing unprecedented levels of design activity. And that is, of course, the early leading edge usually of indications of more robust opportunities and market sizing going forward.

Second question for you, Igor. This is on the logic side. The move to 65 and 45 nanometers are placing a lot of pressure right on the traditional probe cards vendors, given increasing pin counts, pads, pitches and the like. So on that note, could you give us more of a detailed update on the [evalves] with your early fine-pitch logic customers?

That is basically a new revolutionary technology and a new revolutionary product, because what we are doing in logic right now in driving higher parallelism in fine-pitch wire-bonded logic is similar to what we did in '97, '98 with DRAM. I believe this is as important a product that we are working on over the long-term.

We are working with one customer. We are running wafers with our new product. It is a key and very involved from engineering standpoint and applications standpoint installation. But we believe that -- we are happy where we are with this new product.

Okay, great. And last quarter for --

By the way, also, parallelism and pin counts are going up in flip-chip SoC applications. And a significant portion of our flip-chip now revenue will be higher parallelism products, by 2 and customers would like to test by 4. Some of those probe cards have 10,000 springs in them. So all of that is working towards our I think advantage and sweet spot.

Big opportunity there for you guys. And then final question for Ron. What was the level of turns business in the fourth quarter, and what do you expect that to be in here in the March quarter?

I commented that we were in the low 40% range, as we began '06 and we exited about 60% turns. And we don't know precisely how that will play out on a quarter by quarter basis. But what we do know is with the faster cycle time in our factory and the shorter lead-times we are now offering to customers, which is a great positive competitively, that we can operate with higher turns rates. And we will naturally operate with higher turns rates. And our customers don't need to put bookings in place as early as they did in the past to make sure they had capacity opportunities.

So that is changing the dynamic, as we have gone through the year. And we see that dynamic continuing going forward. So I don't have a precise number for you, but they are going to look more like 60s in the future than in the past. That is for sure.

Again, great job on the quarter.

Timothy Arcuri, Citigroup.

A couple of things, Igor, first of all, can you tell us how much of the sequential increase in bookings -- so, whatever bookings number you expect in March, you know, you're not actually guiding to that number. But whatever that number is, how much of that bookings increase will be this large order that you announced today? Will it be almost the entire bookings increase?

We are not guiding bookings for Q1. And we did not size the particular size of that order, other than saying it was a large order. So we do -- let me put it this way. I think the way to think about it is that we certainly expect in the first quarter that we are going to be getting production order traction in a significant way in our NAND products. And this is the first.

Okay, let me --

But that is not the only driver, Tim.

Okay, let me ask you this way. What do you think DRAM revenues will be in the first quarter? Will they be roughly the same kind of high-60s rate they have been the last four quarters? Is that the right way to think about it?

Well, again, we do not attempt to parse the market forecast down to that level. Clearly, DRAM is a cornerstone of our business, and certainly will be in the first quarter. And we see continuing business drivers, as Igor characterized in some detail in his script. So I think we've typically varied in the high 60s to mid-high 70s range for DRAM.

Okay. And then last question I would ask is if you look at your DRAM revenues on a per-gigabyte basis, so if you take the industry's shipments and you look at your DRAM revenues relative to those shipments, do you think -- that number has come down pretty significantly this year. Obviously, there has been some tooling issues with some of the technology transitions.

But when you look at that number in history, is there any reason to think that when the new tooling cycles happen that your DRAM revenues per industry gigabyte being shipped should be any different than what they have been in the past?

The way to look at it is, as I mentioned before, if you look at, say, the total business that one factory will derive from 90 nanometer-based designs, we estimate that the total business will derive from 70 nanometer designs will be 50% larger. And it is about a two-year cycle -- tooling cycle.

So 80 nanometer is on top of that. So that is the way to look at the macro picture. And 90 nanometer was about 50% larger than 110 nanometer, so that is the way we look at it. So that is staying similar.

Now on top of everything of course in DRAM, there is a significant diversification of DRAM revenue. Say, five years ago, it was mainly PC bound. Now, you have very significant portion of it that is consumer applications bound. So that is another way to be looking at that.

Edward White, Lehman Brothers.

In looking at the SoC market, I think that is one of the markets that everyone has been looking for a parallelism. And it has been difficult to achieve technically. Can you talk a little bit about the technical requirements that you have to do to achieve that, the benefits you can give to the customers, some of the customer issues as they are looking at transitioning to high parallelism for SoC?

Yes, well, whenever we get an opportunity to find a problem that is hard enough for anybody else to solve, we love it. That is the way we build the Company. In this case, this kind of the opportunity. What you basically will have to do is to build high pin count probe cards, where pad/pitch requirements will be going down to 40 micron and below, and you will have small pad sizes. And you'll need to hit every pad in the middle, and leave just the probe mark that looks the same on every pin.

In other words, this is a very, very significant electromechanical problem to solve. And say you compare it with NAND Flash applications where customers in the past could afford to design the devices in a way that there is a big pad, and potentially use needle probe cards even for a two/one touchdown probe card, right? This is a completely different classical problem that you need to solve. Once you solve the problem, the market opportunity is immense, in our view.

So it is just precision of probing. And in the future in a few years, you will also see transition to known good die testing in SoC, which will add on top of probing precision the requirement for electrical precision. And that is always a hard problem to solve. If you solve it, you have yourself a good business.

One follow-on question -- on the two-touchdown DRAM probe card Harmony, given the cost of the ceramic substrates, which I understand are more costly with traditional probe cards, how are you looking at manufacturing costs there? Have you been able to find some way around that? Or are there things you can do about that so that you could bring a cost-effective product to the market?

Yes. In this particular case, again, we need to come in with a product that in the end of the day provides very, very significant total cost of ownership advantage to customers so customers, if they are able to do two touchdown probing on a 300 millimeter wafer, that is a huge advantage to customers.

We, from the beginning, when we designed the product, we have target kind of -- targets one course and all the other things. And we work together manufacturing costs and we work on that. There are some ingenious architectural attributes on that probe card that upfront address some of the issues you brought up. But you are right. Those are important issues to address upfront.

Doug Reid, Thomas Weisel Partners.

Hoping you could give a little more color on the revenue guidance. Ron, you mentioned you did not provide full granularity. But perhaps you could just elaborate by geography, by market, where your confidence levels are high regarding the outlook for Q1, where you think there might be some risk? And also, if you could add a little bit of color to the trajectory of known good die, anything there would be helpful.

Yes, Doug. The challenge with a high turns environment is that the money is in the mail, if you will. We don't have as much backlog to look at. So I don't have a great characterization of that, and I use views of design activity and business levels to try to get a sense of that.

So it is difficult to give a more detailed granularity on the revenue numbers. But in general, we certainly as I mentioned a couple of minutes ago, expect to have improving volumes where we are penetrating new markets, such as in Flash. We certainly see strengthening in certain regions, such as in Korea, and see expanding opportunities there as we enter 2007 year.

And in general, our customer base across the board is engaging for 70 nanometer activities and those sorts of things. So where they are ramping earlier with some of our customers in those activities, they will start to ramp up early in the year as well. But I don't have the specific breakdown for you.

On known good die, we expect the growth in the business that was basically the center line of our known good die business in '07 -- in '06, which was for mobile applications, so it is mobile RAM. It is PSRAM, and it is NOR Flash. But also, we are beginning to see -- and by the way, in mobile RAM, you have new entries, as we understand. There will be more capacity shifted to it.

The beauty of mobile RAM is that not only you now sell up to three probe cards for design when you used to sell just one, but the third probe card, which is today majority of our business, is -- there is more sort probe card dollars per bid in mobile RAM. The reason being that the test times are longer for mobile RAM, because there are certain refresh cycles that you need to test for due to low-power nature and all these other things.

Now what we're seeing now in 2007 is we are seeing the beginning of people actually looking at commodity DRAM -- shifting commodity DRAM testing from package to wafer. Now, that is going to be a long-term trend. And these are no longer just kind of technology roadmap discussions. We see some memory makers seriously working on that.

Now, that is something that is not completely -- if indeed, this were to happen, this has not been completely baked in in our growth projections. So that would be an extremely exciting trend for us. The way to think of it is when you do final package test, it is of less value to a semiconductor manufacturer, because you already invested all the money in [huge losses]. You invested all the money in packaging and test.

So it's worth to you some value. If you were to go and do the same test on a wafer before you yielded due to -- after redundancy repair achieved -- before you invested in a sampling test -- and as you know, packages are so expenses -- it is worth a lot more to semiconductor makers. Therefore, the value for us, a final test moving from package and the wafer is beyond just the size of package test market. It is going to be actually the significant amplification.

So we are working on that. We are seeing some real demand out there, starting potentially this year.

Colin McArdle, Needham & Company.

Thanks for taking my question on the Livermore expansion. I was wondering what quarterly revenue level Livermore can support currently, what level you would like to get it to in what timeframe, and how much does it cost?

I can give you a capacity sizing. I mentioned in my comments that we currently have about $125 million per quarter capacity, and that is all in Livermore today. And we can increment that capacity as needed up until the time we get production ramped in Singapore, which will be the second half of 2008, with our current timeline. And that can be done by incrementing certain machine centers where we are reaching capacity and breaking those bottlenecks, if you will.

So the incremental costs of expanding at Livermore are not significant. I.e., you don't have a whole set of costs to that increment, but just breaking bottlenecks in the process.

The other thing I'd mention is we are continuing to learn in our factory improving yields, productivities and throughputs. And that has contributed significantly to our cycle time improvements and improves our cost structure in a very significant way going forward. And that has contributed to our available capacity.

So in part, it is investments we are making to improve capacity. But in no small part, it is also the fact that our factory's improving its performance. And so we have expanded capacity, and that is very low-cost capacity to put in place.

Could you just help me understand how you decided on Singapore as opposed to another low-cost manufacturing center?

FormFactor actually considered this very carefully. And we took time to think about that. We are a maker of some of the most advanced electromechanical products made on earth. I mean, these are very, very complex products. These products increasingly are a part of a system and become a system. So we need certain type of people, personnel to grow the Company.

Such people exist in Japan. It is a difficult place of course to do that. Singapore is a location where government of course encouraged us in very strong tangible ways to expand manufacturing there. So that will, I believe, improve our profitability picture moving forward.

This is a place where you can higher people all over Asia and bring them to Singapore, and Singapore government gives you incentives to do that. This is a place where you have a class to -- therefore to qualified personnel.

It's a supply chain basically capital of Asia. And it takes minutes to go through customs. Where in other places, it may take several days. That is of huge importance for us, because our lead-times are coming down, and we are supplier of custom products for design.

So from that standpoint, and 70 -- you know, huge portion of our revenue is in Asia. So from those standpoints, Singapore was a good location, we believe, for us.

The way we will do it, we will do it somewhat differently. It is not going to be an appendix to the Company as we have done in Japan and will continue to do in Japan. We will encourage people there to be entrepreneurs. I am an entrepreneur. I understand power of entrepreneurship. And we will encourage people there to help them to start potentially new businesses. We will do more R&D there. We will encourage them to do manufacturing in a world-class way. So we are very serious about following through with Singapore.

(OPERATOR INSTRUCTIONS). Mehdi Hosseini, FBR.

I have a couple of questions, first in terms of your operating margin. It seems like in Q4, you had to offset your revenue guidance. But you hit the high end of the operating margin guidance. So my question regarding the March quarter, do you feel comfortable with the margin profile of 20 to 24%? And what is the degree of variation here and considering what happened to Q4?

And also, regarding the DRAM booking, can you give us the color what the mix of DDR2 is as overall DRAM booking in Q4?

In terms of the -- if I understand your question, you are asking about the comfort level with my guidance.

Right, right. Well, I am just looking at the Q4 results. You upsided revenue compared to your guidance. The operating margin came -- the pro forma operating margin came at the high end of your guidance.

Yes, understood. So obviously, there were several factors that I mentioned in brief a few minutes ago that are impacting our guidance. And when you're looking at operating margin line, some of those are fairly well known, such as fringe effects, et cetera.

Another big factor that affects our margin is the revenue level. That has a big impact and the total production coming out of our factory, because of the high fixed-cost nature of it. And that is why you need to range the number and see how that plays out. So I guess that is the best context I could put it in is that it is our best view given the range of revenue we are talking about and expected production levels.

In terms of DRAM bookings, you were asking about DDR2 levels?

Yes, the DDR2 mix of the overall DRAM bookings.

Okay, well, it was in the neighborhood of two-thirds, 70% of our DRAM total on our bookings number.

Okay. And then just going back to the margin profile, with some discretion with your capacity from -- correct me if I am wrong -- from 120 and maybe in your revenue to 170, does that assume that your margin -- operating margin should be peak at 24?

I did not understand. Was it 120 million -- what did you say?

If I understand your capacity plans, you are planning to go from 120 to 170 million without the Singapore facility, correct?

We did not specifically give a 170 number but said we could size capacity as needed until we get to Singapore. And as you heard Igor's comments, we believe that we will be able to grow in '07 above the 25% growth rate of the advanced probe card market but did not specifically peg how we would exit '07.

If that is the kind of capacity ramp, should we also assume that your operating margin should peak around 23, 24%?

As Ron said, I believe, earlier that in 2007 on an annual basis, we believe we can stay on our model of 25% operating margin.

Got you.

Were you talking about a non-GAAP number? Because, yes (multiple speakers) I already commented that that would be 25% for the year.

Yes, I am talking about 25. Otherwise, it would not be consistent with what you had reported.

Right, I commented it would be 25% for the year.

Okay, thank you.

So average that. All right. With that, we thank you all for joining our conference call and look forward to seeing you at the upcoming conferences and on the next quarterly earnings call. Thank you very much.

Thank you.

Ladies and gentlemen, this concludes the presentation. You may now disconnect. Thank you, and have a good day.