The Business of EDA:EDA Business Models

EDA Business Models

Nora:

Do the business models for the EDA vendors vary as much?

Frank:

Well, they vary, but not as much.

Developing the next leading-edge product while supporting your current products is a high-risk venture.

Many large and small companies compete, but at any given time only one of them has the best-in-class product. Small companies are more likely to succeed since they can focus their resources on one product. (There is no cushion—they must sell the new-generation product or go out of business.)

Moreover, vendors do not want to lose the revenue from existing products. In addition, customers are often slow to migrate and want continued support and upgrades. A large vendor has more support staff for existing products. That may limit resources available for risky (but essential) new product development.

Nora:

How easy is it to market a new tool?

Did You Know?

The leading product company in one technology generation will usually NOT be the leader in the next generation. That is an interesting observation that applies for all kinds of businesses, not just EDA.

New EDA Tools

Frank:

It can be difficult for a small developer to sell its EDA tool. Small customers prefer to wait until a large respected user has tried a new tool. Most users want a reference or some success history. With such a reference, an EDA tool vendor has a much better chance to sell a new tool.

However, getting a large customer to try a new tool is also not easy. The customer design engineers (and management) are usually very risk-adverse, since they are paid for on-time success. They have a deadline to meet. They know how to use their present tools. They don't know how well (or if!) the new tool will work.

There is also a substantial learning curve to use a new tool. It can take several months to learn a new EDA tool. The engineers have to try it out and compare it to other new or existing tools. (Internal development groups face the same hurdle, by the way.) The company has to invest time, money, and staff.

Therefore, an EDA tool vendor must show that their tool gives a large (maybe 5X-10X) improvement. This is usually in run time, performance, design size, or accuracy. Sometimes customers will try a tool simply because there were no other options available to meet their project deadline.

At the same time, there is a symbiotic relationship between the EDA vendors and the users. The large users need the latest tools for their innovative designs. Most customers need new tools frequently to keep up with increased chip complexity, speed, and new issues. The vendors need the users to check out the tool and encourage other buyers.

Nora:

Well, marketing a new tool sounds like an interesting challenge. How will Sandbox grow?

Licensing Models

Frank:

A great deal of company revenue depends on the licensing model it uses for revenue. The company can sell the tool, license it for a fixed or variable time, license it to a customer company or individual user by calendar time or usage, etc. The choice of a competitive and profitable license model can make or break the company. Licenses are often negotiated on a per company basis.

Yearly maintenance charges (often 15% of the sales price) for updates and bug fixes also contribute significantly to the revenue and cash flow picture.

Mergers and Acquisitions

Frank:

In order to grow, a small tool developer must either be acquired or merge with other small vendors. Acquisition is usually by one of the three or four large EDA tool vendors.

Alternatively, the small vendor needs to develop a series of successful new tools. Historically, as I said, the chances of the same group developing another "star" tool are very low.

For the large tool vendor or OEM, the problem is similar. It is hard to maintain a leadership role. Acquisition of small promising startups is one way that companies catch the "next wave."

However, the integration of one company into another is always tricky. Different cultures, expectations, and management styles make integration from mergers and acquisitions difficult. There is often the issue of a choice between competing internal products and acquired external products.

Many promising EDA tools (and their developers) disappear after an acquisition or merger. There are some ways to sweeten a merger. However, there is no way to ensure that key technical people will not leave.

An example of such a merger is when Synopsys, a large EDA company, acquired Avant!. The Synopsys culture runs on consensus, encouraging employees to speak out. At Avant!, employees reportedly were expected to do as they were told.

In addition, some Avant! tools were complementary to Synopsys' tools and others were competitive. There were some hard choices on which tools to keep and which to drop.

Another example was when Valid Logic merged with Cadence. Cadence attempted to alternate management layers. For cultural reasons, this did not work very well. Groups often form a strong sense of identity, which can conflict with such a shakeup.

Nora:

Perhaps "explosive growth" doesn't apply to the EDA business. Are there other business models?

Application Service Provider Model

Frank:

Yes, a few others come to mind. The Internet has led to another business model—that of the Application Service Provider (ASP). In this model, the EDA vendor hosts a tool or service on the Internet. The vendor sells the tool or service on a usage basis. (Either by time or some program completion event.)

The user does not get the tool source code or use of the tool on their own machine. The ASP vendor provides a server on the Internet.

Nora:

I imagine that should be a major growth area?

Frank:

Possibly—although remember that most EDA involves very large, secure data files and transfers. The Internet is not so good for those.

Design Services Business

Some EDA companies also try to expand their business models with Design Services divisions. They provide designers or consultants to help an electronic product company execute a design. They will help with or do the whole design. Design services can complement the EDA tool side of the business, providing a testing ground and reference.

Design services, however, do not have the same leverage as software EDA products. A small software staff can create a product that generates a lot of revenue. The revenue is not in direct proportion to the cost. That's leverage—creating a lot from a little. To grow in design services, you have to increase the staff in direct proportion to the revenue. If sales fall off, you still have a large staff to feed.

Other EDA companies have acquired and sell re-usable design blocks, called Intellectual Property (IP). Re-use of predesigned blocks saves a lot of design time as chips get more complex. Therefore, this practice has been growing slowly but steadily. The IP business has too many pros and cons to go into here. (The reader can learn more about IP in Appendix F.)

Some companies form or join standard groups trying to promote their tool approach as an industry standard. Others form alliances with vendors of complementary tools to offer a more complete tool suite for their customers.

Nora:

Thanks, Frank. There really are quite a few ways to make money in this business.

Frank:

Yes, but remember the EDA business is very small compared to its customers.

Nora:

How do they compare?