2009 was quite a tumultuous year for the EDA industry. However, there are signs that there is some economic light at the end of the tunnel. Revenue improved on a sequential basis in the third quarter of 2009, ending the longest string of successive declines since at least 1997, according to the EDA Consortium (EDAC) trade group.

EDA revenue for the third quarter of last year totaled nearly $1.17 billion, up 3.8 percent compared to the second quarter of the year, according to EDAC's latest report. But revenue continued to decline on a year-to-year basis, down 7.2 percent compared to the third quarter of 2008. EDAC's four-quarter moving average also declined 13.1 percent.

EE Times presents a discussion with EDA pundits Walden Rhines, EDAC chair and chairman and CEO of Mentor Graphics Corp., Gary Smith, principal of Gary Smith EDA, and Joseph Borel, former executive vice president in central research and development at STMicroelectronics NV and now of the JB-R&D consulting company.

Note that each executive was asked the same questions. Walden C. Rhines, EDAC

Walden Rhines EDAC Chair

EE Times: Will the EDA market stabilize in 2010? What will be the positive trends and negative trends?

Walden C. Rhines: I believe the market will stabilize. The last two quarters of EDAC results have shown decreasing declines. We are seeing positive forecasts for the overall electronics industry in the upper-single digits. IC Insights projects 7 percent electronics growth while VLSI Research is projecting nearly 9 percent growth in 2010. The markets with the highest semiconductor consumption, personal computers and cell phones, both seem to be recovering nicely and this is helping drive the overall demand for semiconductors into the mid-teens, according to many of the leading analysts. Semiconductor market watchers are projecting growth rates that will drive semiconductor consumption higher than the last peak in 2008. Fortunately for EDA, both the computer and wireless markets remain highly competitive and will continue to use design to enable both cost reduction and differentiation goals.

Beyond the macro-environment, a portion of the revenue shock in 2009 can be attributed to individual company business model changes. Those changes have been made, the industry has absorbed the impact of those decisions, and EDA can move forward.

EE Times: Is it likely that the EDA pickup in revenues may not be until well in 2010, or even in 2011, when customers are certain that their markets have recovered?

Rhines: EDA growth is most tightly correlated to semiconductor research and development investments. The strongest relationship tends to include a delay of about one year. Research and development investments were impacted in this downturn, but the decline in R&D expense was much milder than we experienced in 2001. While the data for 2009 is not yet available, we are starting to see companies adding to their R&D staff, with considerably fewer companies continuing to cut. TSMC's announcement on Jan. 14 that they plan to recruit 3,000 engineers is an encouraging sign.

The fabless companies continue to outpace the overall market with 15 of the top 50 fabless companies posting positive growth for the year. Even in 2009, during one of the worst recessions, a half-dozen leading semiconductor companies enjoyed positive growth rates.

It is reasonable for companies to remain somewhat cautious, but companies will need to balance caution with their own competitive environment. High technology companies are rarely known to "save" their way into prosperity - they invest, especially coming out of recessions so that they have the products in the pipeline to exploit during the next upturn.

EE Times: Should we expect more consolidations this year?

Rhines: Consolidation in EDA is something of a broadly accepted myth. The "big three" EDA companies (Synopsys, Mentor Graphics and Cadence) have a 73-percent combined market share as of Q3 2009. That market share is within a few percentage points from where it was in 1998. It is in fact the average share that the "big three" have enjoyed since that time — a share that cycles slowly between 70-80 percent.

In most years, new EDA start-ups roughly offset companies that are acquired or exit the business. It is difficult to track exact numbers, but Mentor Graphics has identified 27 rounds of venture funding into EDA companies in 2008. And, as of Q3, we have already identified 15 rounds of funding into EDA start-ups in 2009. While the numbers of rounds and perhaps the dollars invested are down, they have not vanished. There are plenty of design problems looking for solutions and innovative engineers from both large and small companies with the opportunity to contribute.

EE Times: Will big EDA companies keep going or change course?

Rhines: EDA is in reality a collection of distinct product "sub-markets". The leading vendors tend to have large shares of some of the larger established sub-markets. That is often where most of the competition is. Unfortunately, the larger established submarkets, once established, rarely grow more than a few percentage points annually. For example: according to EDAC data, the RTL simulation market has grown at an 8 year CAGR of 1.3 percent; the analog/mixed signal and RF simulation market has grown at an 8 year CAGR of 2.1 percent; PCB layout at 1.2 percent.

All significant EDA growth has always come as a result of the introduction of new design methodologies. DFM/RET tools didn't exist in year 2000 — today the DFM/RET related tools have revenue of $250 million and constitute over 36 percent of the entire EDA license and maintenance growth since year 2000. Likewise, growth in the ESL market has a long-term CAGR of 14 percent and alone accounts for 20 percent of the total growth. Combined that is nearly 60 percent of the total EDA market growth for the decade so far.

Since growth requires new markets, then for serious growth to occur the big companies will have to evolve to solve the new problems in areas such as system design, high-level design, and embedded software.

See also: http://www.eetimes.com/222300878 Gary Smith, Gary Smith EDA

Gary Smith Principal Analyst, Gary Smith EDA

EE Times: Will the EDA market stabilize in 2010? What will be the positive trends and negative trends?

Gary Smith: Yes, the Cadence accounting problems will be over and the semiconductor recovery will drive growth. The negative trend will be the continuation of the Upper Mainstream users dropping out of RTL & Below design (implementation really).

EE Times: Is it likely that the EDA pickup in revenues may not be until well in 2010, or even in 2011, when customers are certain that their markets have recovered?

Smith: I don't think so. Pent up demand has been evident for a quarter now, and the move to 32nm demands new tools.

EE Times: Should we expect more consolidations this year?

Smith: Yes, I am expecting a roll up in the middle. The likely candidates are SpringSoft, Apache and Atrenta acquiring other smaller companies. Usually, in these inflection points a new Big EDA company is formed. This time, I think it will come from the mid-size companies as it did with the Cadence roll-up during the RTL inflection point.

EE Times: Will big EDA companies (Cadence, Mentor, Synopsys, Magma) keep going or change course?

Smith: With the exception of Mentor, they will change course or fade away. Mentor is well-positioned for the future, the rest are not. Joseph Borel, Catrene European EDA Roadmap

Joseph Borel JB-R&D EDA consulting company

Borel, now of the JB-R&D consulting company, co-wrote the 2009 European Roadmap for design automation in semiconductor products.

EE Times: Will the EDA market stabilize in 2010?

Joseph Borel: I don't see any reasons why EDA market will stabilize in 2010 because EDA market by construction is not consolidated and that's a prerequisite for its stabilization at the required level.

What I mean is that there should BE NO stabilization but rather a significant move towards EDA industry standardization and more sharing of complementary industrial offers (e g new system level solutions to address design needs from system houses functional specifications to support their ramping up in the semiconductor market share).

The implementation way is the common adoption of an International EDA Roadmap, as there is an International SIA technology roadmap. The European CATRENE EDA Roadmap could play a role as a starting template of the work to be done.

I proposed a statement on this topic at the latest Design Automation Conference in San Francisco, Calif., that should be seriously and positively analyzed by the current EDA players, for their own benefit (including system houses, semiconductor companies, EDA companies and EDA startups).

EE Times: What will be the positive trends and negative trends?

Borel: A positive trend would be more dialog between EDA customers, including from one side customers —system houses and semiconductor companies— and from the other side suppliers —EDA vendors and EDA startups including tools and libraries with an emphasis on System Level Design and eSW with a specific attention to application using low cost and low power FPGA solutions.

There would a negative trend if that does not happen!

EE Times: Is it likely that the EDA pickup in revenues may not be until well in 2010, or even in 2011, when customers are certain that their markets have recovered?

Borel: Real pickup will occur only when EDA suppliers understand the above, their present business being still in the noise level!

EE Times: Should we expect more consolidations this year?

Borel: It will take at least five years to reach the above targeted level together with making available in parallel the new EDA tools needed for the 3D integration that could well see their industrial move slowing down the 3D product developments on the market.

A new DDA (Device Design Automation) level will become mandatory in this business. This is related to 3D Integration in terms of "Multiphysics Device Simulations" to study process variability and its impact on devices at the nanoscale level and then circuits variability

EE Times: Will big EDA companies (Cadence, Mentor, Synopsys, Magma) keep going or change course?

Borel: In my opinion, they will need to change course to cooperate more closely and share a much bigger cake than what it is today!

See also:
Ex-ST exec calls for EDA to take strategic view EDA from "soft business to competitive business"