PDF Offers Value, Finds Scaling Difficult

I reiterate – not much has been happening in the EDA space. With Cadence having crashed earlier this month, the industry continues towards its bleak financial future, unless it takes necessary decisive steps. (Read more on Future of EDA and Future of EDA Addendum).

Yield is one of the big problems facing the EDA industry – as chips grow denser, yields suffer. EDA has done very little so far to impact yield. Design for Yield remains, by and large, a distant dream. A real pity!

PDF Solutions (NASDAQ: PDFS) is a small-cap company and a leading provider of yield improvement technologies and services, a company worth discussing in this context.

PDF’s value proposition is to improve yield by optimizing design and manufacturing interactions in the semiconductor manufacturing life cycle from initial process ramps to mature mass manufacturing operations. It has hardware (test chips with core modeling capabilities) and software (manages yield and detects faults) solutions for yield enhancement.

Its gain share business model model allows the company to participate in the “volume” semiconductor business, as opposed to EDA companies that are tied to design starts. In other words, PDF gets rewarded for the yield enhancement it delivers for customers.

Despite all this, its stock price crashed following the announcement of its Q4 results on February 7, 2008.

The company reported a year on year growth of 30% from $19 million to $24.6 million for Q4. Sequentially, the revenue remained flat, growing at a marginal 2% from $24.1 million in Q3. The revenue was below market expectations of $37.0 million.

PDF’s diluted EPS on a Non-GAAP basis was at $0.21 compared to Q3 reported $0.19 and previous year’s $0.08. The EPS was marginally higher than market expectations of $0.20. During Q4, PDF had repurchased 201,300 shares from the open market at a weighted average cost of $7.57 – a repurchase value of $1.5 million.

For the year 2007, PDF reported annual revenues of $94.5 million (markets expected $95.0 million) compared to 2006 revenues of $76.2 million – a 24% increase. Their EPS of $0.71 was marginally above market expectations of $0.70. In the year 2006, the earnings were at $0.46 per share.

Segment wise, DFM Yield Solutions and Services contributed 71.8% of the company’s revenue compared to 60.8% the previous year. Q4 also saw Gain Share contribute 26.5% to revenue (second largest revenue share from the segment in the history of the company) up from 25.7% in the previous year. Revenue from Software Licenses fell from 13.5% in the previous year to 1.7%.

For the year 2007, contribution to revenue from DFM Yield Solutions and Services registered a substantial increase over the previous year’s 59.6% to 67.5%. Revenue from Software Licenses reduced from 14.1% to 7% and for Gain Share from 26.3% to 25.5%.

PDF announced its projections for the next fiscal quarter to be in the range of $23.5 million to $24.5 million. Its revenues have been in the same range for the last three quarters suggesting a flattening of revenue growth. Analysts were expecting revenues of $25.6 million for the period. It projected Gain Share revenue in the range of $5.3 million to $5.8 million.

On a Non GAAP basis, net income (loss) for the quarter is expected to be $0.08 to $0.10 per diluted share compared to market expectations of $0.14.

The share has been hitting new lows since the announcement. On February 8, 2008, the share fell to $5.57 from the previous day close of $8.26 – a 33% market erosion. In fact, the stock is at its lowest in the last five years. Currently trading at $5.66, its market capitalization has reduced to $158.70 million. The stock is also seeing high volumes being traded. Compared to the average volume of 127,000 shares traded in the last three months, over 2 million shares were traded on 8th February alone.

I have long been of the opinion that a marriage between PDF Solutions and a major EDA vendor (Cadence or Synopsys) would be good for both parties.

PDF is essentially a highly specialized consulting shop. It has the industry’s best talent and core competency in yield optimization.

However, it absolutely does not know how to build products.

Cadence and Synopsys lack the yield core competency, but they do know how to bring software products to market.

They need each other, PDF and the EDA guys.

PDF has another problem. To scale, they need to figure out a way of “selling” their highly specialized offering to a larger group of customers, as well as, “delivering” on the promise. This has not been an easy challenge for the company, and will continue to limit the company’s growth potential.

Instead, the fantastic knowledge that they have accumulated needs to be encapsulated in software “products”, and I continue to be of the opinion that marrying an EDA vendor would facilitate that process.

And at $160 Million, Cadence or Synopsys would likely be very happy to embrace PDF with a premium. Its gain share business model is of great interest to both, to get out of the design start rut.

Mike Fister and Aart De Geus could observe the Microsoft-Yahoo! saga for some lessons.