First of three parts: EDA’s history of acquiring startups has worked better than most other industries, but it’s not always apparent—and it may not last.
By Ed Sperling
The rule of thumb for mergers and acquisitions is that the majority will fail. So why, despite concerns about big companies buying up the tools of startups, does EDA’s track record look so good?
There are a number of answers that are unique to the EDA industry:
There are technical reasons, as well. “EDA is done through interfaces versus integrated products,” said Jim Hogan, a long-time EDA venture capitalist. “That’s one of the reasons why Open Access was so important. It enabled integration and made acquisitions easier. But the industry has always done a good job of integration and formats. Look at SPICE and SystemC. Those are generally open. That allows you to buy a company like Altos, which was doing about $10 million in revenue and had done the interface, drop it into the Cadence channel, and suddenly you have more feet on the street and can grow that revenue.”
This wasn’t always apparent to EDA companies. Jack Harding, president and CEO of eSilicon, breaks the often-frenetic history of EDA acquisitions into three segments. “In the early days, even without integration skills, the entire organization knew how to integrate tools to make it work. That’s the hard part. The administration and G&A part is easier. If you look at failed acquisitions, the products were not integrated at the rank-and-file level.”
That was followed by the second phase, where executives had learned how to buy and integrate companies. And in the third phase, which includes the last 10 years, he said there are companies being built for acquisition by the Big Three EDA companies.
But there also are questions about whether EDA will prove to be a victim of its own success. The market simply has been overfished. With fewer startups being created, not to mention less VC funding going into EDA, it’s harder to find acquisitions. And with the bulk of innovation still coming from startups, that poses some long-term issues for the largest EDA companies. A quick look at the history of EDA shows that Big Three EDA companies have garnered a consistent share of revenue for the past 20 years—with a steady stream of acquisitions. What will happen without that level of acquisitions is anyone’s guess.
The case for acquisitions
There are four main reasons why EDA companies buy other companies. One is to bolster an existing area of strength—segments where a company already is a leader in a particular market. A second is to expand into a new or adjacent market where there is a growth opportunity. Third, it’s often cheaper and quicker to acquire than to build. And fourth, by buying you whittle down the competition.
“The number one question we ask when we acquire a company is whether we can add value to the acquisition,” said Wally Rhines, chairman and CEO of Mentor Graphics. “The number two question is whether it builds on our strengths. If we’re already number one, it may allow us to concentrate more on a solution. If we’re the leader, we have brand recognition and a customer base and we can create more value for the customer. That also means management has more time to be flexible with the acquisition.”
Synopsys, which has been rolling up companies of all sizes in areas such as IP, virtual prototyping, and more recently emulation, views acquisitions in terms of a value proposition. “There are two fundamental ideas,” said Aart de Geus, Synopsys’ chairman and co-CEO. “One is what gets created for the customer. The second, is how you deliver that value, which includes the culture and the execution machines.”
De Geus noted that there have been a number of books written on acquisitions gone wrong. “The problem is that so many of them have been done out of a position of strategy of weakness—short-term answers for how to fix a present problem. That tends to overload the boats, rather than understanding that what you need to do is switch direction. But if you look forward with a more objective and realistic view, you find you can assemble systems with different pedigrees. What makes this work is when you cherish the strength of the company you’re acquiring and you’re willing to align the cultures.”
Synopsys, like all of the Big Three, has had lots of practice at this. Last year it acquired nine companies, stretching its own internal team to the point of exhaustion.
Cadence unveiled its own acquisition strategy over the past couple years, moving from an unknown in the IP market to the No. 4 position. But Cadence is no stranger to acquisitions, either. In the early days of EDA, it created an entire industry through acquisitions. The most recent round includes the acquisitions of Denali, Tensilica and Cosmic Circuits.
“With every acquisition, you need to debate it internally and take your time,” said Lip-Bu Tan, president and CEO of Cadence. “You need to be very clear why you acquire another company. You have to enhance the solution for the customer. It’s not just about revenue.”
Picking the right pockets
Acquisitions work two ways. Tan noted that Cadence worked with Tensilica for three years before acquiring the company. He said that deal hinged on being able to keep a couple hundred people that worked for Tensilica.
On the flip side, Apache Design spent months looking for a partner to grow its sales and extend its product line, rather than just seeking out the highest bidder. Because most of the companies being acquired are private startups with the founders deeply involved in the company—rather than public companies that have to serve their stockholders—they can afford to be choosy. For a deal to work well it has to serve both sides for many years.
“Every merger and acquisition is different,” said Andrew Yang, president of Apache Design Inc. “What we were concerned about was first of all that there wasn’t overlap, so it could be expansion based, and second, we wanted to retain our expertise and our core people. That’s the only way you can make sure it does not impact customers, products and people. We kept our sales channel and support independent so it did not do any harm to customers. And that was our number one priority—do no harm.”
This was somewhat unique, because most acquisitions tend to leverage a larger sales channel of the company making the acquisition. In fact, one of the most difficult jobs in EDA company integration is integrating those channels. Yang said this acquisition was engineered from the company being acquired up, rather than the other way around.
“Our strategy was always chip to package to board, and eventually to system,” he said. “ANSYS was the leader in chip-aware systems and they wanted to play in mobile. It turned out to be a great match.”
A growing number of companies also are built from the ground up for acquisition by specific companies, as well. Typically those include some former employees or business partners of those companies, who can identify a clear need and a strategy for integrating within the acquiring company even at the startup’s inception.
Next In Part II: Integrating and aligning companies, and metrics for success.