EDA Industry: Consolidation Remains a Priority

As per a recent report by the EDA Consortium, the industry’s revenues have declined 10.7% over the year in the first quarter of 2009 to $1.192 billion after registering single-digit growth in 2008 and double-digit growth in 2007. Computer-Aided Engineering, the largest product category, saw revenues fall 19% over the year in the first quarter to $427.9 million, while IC Physical Design & Verification revenue, the second-largest category, fell 5% over the year to $302.2 million. Asia-Pacific was the only region in which there was growth. Another report points to declining semiconductor industry capacity. Overall installed capacity for 2009 is expected to decline by 3% to 15 million wafers per month. With such statistics, one can’t expect any of the four key EDA players to improve their financials in the near future if they continue to treat this as business as usual.

The largest player in the industry, Synopsys (NASDAQ:SNPS), reported almost flat revenues. Q3 revenues of $345.2 million met the Street’s expectations and grew 0.3% over the year. EPS of $0.47 exceeded the Street’s expected $0.41 and grew 7% over the year.

License revenue accounted for 88% of revenue while services and maintenance grew 22% over the year to account for 12% of total revenue.

Synopsys is making advances in its product offerings. The company recently introduced a new physical verification capability in their Galaxy Implementation Platform with the IC Validator, which will let customers diagnose and fix issues during place and route before final signout, thus reducing the number of schedule delays. Earlier this year, they shipped a new Galaxy release with more core features and better low-power performance. They are able to demonstrate to their customers higher efficiency through the product. At the recent Design Automation Conference, Synopsys received several accolades from their customers for Galaxy and the in-design capability integrating the IC Compiler with synthesis, sign off, and physical verification. They are launching similar upgrades within Verification.

While Synopsys might still not be looking at consolidation within the industry, they are at least finding partners within the value chain. They recently announced collaboration with ARM and the IBM/Samsung chartered common platform, through which they are developing a comprehensive solution for high-volume mobile applications. The product at 32 nanometers will result in lower costs and faster time to market.

Synopsys’ recent acquisition of MIPS’ analog group has added a catalog of analog IP blocks and nearly 450 IP designers to Synopsys’ portfolio. Customers expect the acquisition to help Synopsys strengthen their current IP product offering and become a leader in the segment. This is a high-margin business, and it’s a good area on which to focus.

Synopsys is projecting Q4 revenues of $335 million to $343 million, falling short of the market’s estimated $348 million. EPS is projected at $0.29 to $0.33 and is marginally shy of the market’s expected $0.34. For the year, these expectations translate to revenue of $1.357 billion to $1.365 billion with EPS of $1.71 to $1.75.

The stock is trading at $21.52 with a market capitalization of $3.1 billion.

Cadence (NASDAQ:CDNS) has had a tough time, not only because of the recession, but also because of changes in its top management. The company has been struggling with leadership ever since Joe Costello moved on. With Lip Bu Tan in the driver’s seat, I am hopeful that they will stage a recovery. I recently sat down with Lip Bu, and caught up with him on both the state of the union and his plans for turning around Cadence. Lip Bu is working hard to improve employee morale and customer confidence, but he believes that it will take about three years to undo the damage done to Cadence’s culture by the two previous CEOs. While Ray Bingham’s quarterly earnings-oriented management took a toll on Cadence’s R&D, Mike Fister’s arrogant and disengaged non-leadership and milking of the Cadence coffers have left the company gasping. The board had been asleep at the switches, and now Lip Bu Tan, one of the members of that sleeping board, has stepped in to take the reins and effect a turnaround.

It is too early to see the results of Lip Bu’s leadership, but I liked what I heard in terms of general direction. Recently reported Q2 revenues of $209 million were short of market’s expected $216 million and 32% lower than the previous year’s revenues. However, loss of $0.05 per share was better than the expected loss of $0.13 per share. By segment, product revenues fell 42% over the year to $102 million, services revenues fell 18% over the year to $28 million and maintenance revenues fell 19% over the year to $80 million.

Cadence has two strong franchises in Analog & Mixed Signal design sold in the form of the Virtuoso platform and Verification, which they acquired as part of the Verisity acquisition. And they clearly dominate these fields. Recently, Fujitsu Microelectronics Solutions signed up to incorporate the Virtuoso Multi-Mode Simulation and the AMS designer suite of verification software for Fujitsu’s most difficult mixed-signal designs. They also won contracts from the Japan Aerospace Exploration Agency, Toshiba and National Semiconductor for their Virtuoso range.

Cadence is continuing to expand their offering and recently unveiled its first unified TLM-driven design and verification solution and methodology, which will enable SoC designers benefit from its transaction-level modeling.

Cadence is projecting revenues Q3 revenues of $210 million to $220 million with net loss in the range of $0.01 to a profit of $0.01 per share. For the full year, they are projecting revenues of $830 million to $870 million with net loss of $0.20 to $0.08 per share.

The stock is trading at $5.94 with a market capitalization of $1.6 billion.

Mentor’s (NASDAQ:MENT) performance surprised many in the market. Revenues of $182.6 million were flat over the previous year’s $182.4 million and managed to exceed analysts’ expected $165.8 million target. EPS of $0.02 was also significantly higher than the loss of $0.02 per share earned a year ago and compared with the market’s expected loss of $0.10 per share.

After several consecutive quarters of a shrinking customer base, this quarter, this quarter the number of new customers grew 10% and bookings grew 15%. Bookings grew across all segments and were led by Design to Silicon, which houses the Calibre product, reporting growth of 45% over the year.

Mentor attributes their current success to their product portfolio, especially the Calibre software, which has become an industry standard for physical verification. They recently expanded their offering under the TSMC Reference Flow 10.0 to support advanced functional verification for complex ICs and tighter integration with the Calibre physical verification and DFM platforms. The improved product also addresses low-power design with Mentor tools for functional verification, IC implementation, and IC testing.

In addition to DFM, Mentor is now able to offer complete solutions in the DFT category. A complete DFT solution will require both scan and BIST or built-in-self-test offering. With the LogicVision acquisition announced earlier this year, Mentor has consolidated their leading Automated Test Pattern Generation (ATPG) and embedded test pattern compression technology with BIST products to provide customers with integrated solutions to address the complexity of silicon test. It is definitely a compelling reason for customers to focus their testing needs on Mentor, and the LogicVision acquisition was a smart move. Test costs are becoming a key issue for chip vendors, making them look at ‘testers on chip’, which is what LogicVision offers.

Mentor also recently launched their Android and Linux strategy, which includes the acquisition of Embedded Alley Solutions, a leader in Android and Linux development systems. They plan to combine their real-time operating system (RTOS), tools and middleware with Embedded Alley’s products to be able to provide device manufacturers with software needed throughout the product development cycle. They are looking at addressing demand from products based on Google’s Android platform and Linux, such as smartphones.

Analysts expect the coming year to be good for Mentor as many of the company’s three-year contracts are up for renewal and are expected to be renewed at a bigger size. Analysts peg the increase in contract size at 10%.

Mentor projects revenues of $183 million with an EPS of $0.01. Analysts were expecting revenues of $180 million with a loss of $0.02 per share.

Mentor’s balance sheet, however, carries a large debt of $220 million because of their recent acquisitions, which could bring down the company’s valuation in the event of a potential merger deal.

Mentor’s stock is trading at $8.93, taking its market capitalization to $850 million.

Magma’s Q1 revenues fell 37% over the year to $28.8 million with EPS of $0.03 for the quarter.

The company has remained strong in digital implementation by expanding the features within Talus. During the Design Automation Conference, Magma, along with TSMC, announced the inclusion of Talus in TSMC’s Reference Flow 10.0, which offers common customers differentiated design and process technologies to improve power, performance and design for 28-nanometer ICs. They recently announced that their customers passed the 50-tapeout mark for chips designed at 45 nanometers or smaller geometries using Talus netlist-to-GDSII design implementation system – “more than with any other EDA supplier’s implementation system.”

Magma has also expanded their presence in the memory market by releasing a new version of the FineSlip Pro, their fast SPICE engine that can now scale on multiple CPIs. They estimate the circuit simulation market at $325 million, and they are the technology leaders in the segment. The company also managed to bag NVIDIA’s account from Synopsys for Magma’s Quartz product; NVIDIA will deploy Quartz as its primary design rule checker for chips at 40-nanometer and smaller.

Within analog implementation, Magma’s Titan mixed signal platform now includes analog simulation environment and schematic driven layout capabilities, through which they were able to win customers from Cadence, the market leader in the space.

Magma is projecting Q2 revenues of $28.5 million $29.0 million with EPS of $0.00 to $0.02. While they maintained their annual revenue projections, they are raising EPS outlook for the year from $0.08 to $0.11.

Magma’s stock is currently trading at low levels of $1.40 with a market cap of $68 million.

PDF Solutions, (NASDAQ:PDFS), the leading provider of yield improvement technologies, saw Q2 revenues fall 6% sequentially and 55% over the year to $9.6 million. Gain share revenue of $2.3 million fell 4% over the quarter and 60% over the year. Net loss per share of $0.25 was marginally lower than $0.28 a quarter ago and significantly higher than $0.07 a year ago.
 
PDF is focusing on generating more revenue per client by expanding the process life cycle solution and securing longer gain share turns and controlling the customer’s costs. Their yield to aware FDC solution is already gaining market share, and two clients engaged PDF for full deployment under contracts that can generate revenues in excess of $10 million each.

Besides increasing account penetration with their product range, they are also increasing future revenue potential from gain share contracts. In Q2, they signed a 45-nanometer integrated yield ramp with a logic client and converted the 32-nanometer LOA signed in the earlier quarter to a full contract.

Also, as part of their go-to-market strategy to reduce selling costs, PDF is leveraging their relationships with process licensers to reach their client licensee fabs. But, as we saw earlier, the company’s situation remains “precarious,” and I maintain my view that PDF needs to be acquired by Synopsys if it wishes to grow.

PDF’s stock is trading at $2.82 with a market capitalization of $66 million.

I still believe that the future of the failing EDA industry lies in consolidation. If Cadence were to use their cash balance of $562.4 million to acquire Magma, that would really shake the market up. Synopsys’ P&L is the strongest in the market by far, and the company is performing better than the rest of its peers. Cadence has the greatest incentive to lead a consolidation effort over the next two to three years as part of their turnaround effort.

In terms of products, Cadence leads with their analog and verification offerings while Synopsys leads in digital. If Magma were to join Cadence, they would bring these strong digital capabilities with them. Magma recently took the NVIDIA account from Synopsys. Similarly, if Mentor joins Cadence, the latter’s DFT and DFM portfolio would be complete. Add Mentor’s $182 million and Magma’s $29 million to Cadence’s $209 million, and we surely have a new market leader.

Cadence’s earlier attempts to purchase Mentor failed. But under Lip Bu’s more humble, relationship-oriented leadership style, this time, the merger might just work.

[Full disclosure: Cadence and LogicVision have been significant clients of mine.]