New Directions For EDA

DAC is over and everyone is asking – what was the theme this year? It is sometimes difficult to make such a determination because quite often the theme has been there for some time, but suddenly appears more obvious than it did in the past. Some years it is a new product or class of products.

The theme also can remain hidden, or disguised. As an example, people have been talking about the for a few years now. Why is it important? Because some of the tradeoffs are different than they would be for the types of devices that the EDA industry has traditionally targeted. A smart phone is all about squeezing in as much functionality as possible while maintaining an acceptable power draw. The IoT, on the other hand, is about minimal functionality necessary while removing every possible Joule of energy usage for a given task.

The IoT does not care about thermal in terms of heat generated, although it may be concerned with heat coming from external sources. Automotive is another interesting example, where device longevity is an important consideration, something not thought about for a smartphone that will be replaced in two years. The tradeoffs being made are different for these emerging markets, and with those come the opportunity to do things differently.

At Gary Smith’s talk that kicked off DAC, he postulated about why EDA needs to move into different markets in order to expand. He mentioned areas such as mechanical, embedded software and applications software. Smith gave the reasons why these are being looked at as simply a way to expand the market opportunity for EDA and increase revenue, but there is a much more fundamental reason why EDA companies are investing outside of the traditional EDA market and flow.

Consider Mentor Graphics for a moment. The company started to make investments in other areas many years ago. It entered the embedded software market almost 20 years ago and automotive about a decade ago. In order to enter each space it had to buy one or more companies in those spaces. Did it buy those companies directly for the revenue stream? No, it bought them in order to build up the technology, the credibility and to learn how to do business with those industries.

Automotive does not yet show up in the Mentor annual reports for a top-income category, but it does show up as an area that is growing rapidly, even if from a small base. What would stop the other EDA companies from copying Mentor and also going after the same market? A lack of available companies with the necessary credentials to get them into those markets and many years to become accepted by those markets is one reason. Mentor has acted as an aggregator for pieces of technology and made them into a complete vertical solution.

Mentor has such a large lead that would probably be impossible to replicate. Upon entering the DAC show floor this year, and in fact last year, one of the first things that became obvious is the large area dedicated to automotive. This is an important vertical and Mentor has decided that it wants to own that vertical, and has been working on it for years. It is beginning to gain traction and it may not be too many years before Mentor is viewed as an automotive tools company.

This becomes a game changer, not because it is a different market but because it is a market that is untainted by the past business practices of EDA. Bundling, all you can eat deals, price reduction to the point where you are trying to kill a market are things that we have seen in the past when too many EDA companies were competing for the same few customers. Vertical markets are defensible and thus will have different pricing dynamics.

Another example is the recent purchases by Synopsys in the applications software space, including Coverity and Codenomicon and assets from Quotium. Gary Smith sees this as Synopsys attempting to get into the embedded software space and to compete with Microsoft and Oracle. This is highly unlikely. Synopsys has identified a vertical market it is going after and is building a defensible moat around that.

It may well be that this is seated in software security and could tie in with some of Synopsys’ IP developments so it can create a vertical line of products targeting a specific industry that will not be addressable by others, even with the necessary IP. It would have complete vertical products and this does fit into the company’s notion of building subsystems for people rather than just IP building blocks.

Why would EDA be interested in mechanical—a market that is smaller than EDA, has many incumbent companies in an industry that has existed longer than EDA, and according to Smith, is struggling? We only need to look to Ansys for the answer to that. It was a mechanical company that bought into EDA when it acquired Apache.

Ansys did not do it to get into EDA. It bought Apache to own a vertical that had not been identified and tapped by either community. It was after the thermal market and needed technology from the semiconductor side, the packaging side and other technologies such as airflow. Its recent acquisition of Gear, a data analytics company, may well fit into that new vertical it wants to own. There are rumors floating around that Ansys is after other power companies, as well.

Both Mentor and Synopsys have been making a number of thrusts into vertical markets, and IoT is clearly the latest to surface. Did Mentor buy Tanner for this reason – quite possibly, but this is just a technology investment and not something that builds any defensibility at the moment. Nobody appears to have discovered a way yet to make this a market with a moat, but I am sure the best minds in the big three are working on it.

EDA is not getting out of EDA, nor is it buying into other businesses. It is bringing together the technologies necessary to construct solutions and doing it in ways that will make it difficult for the others to follow. Long term, this looks like a promising path to increasing their revenues, even if core EDA continues to get no respect. Core EDA is a technology, it is not a solution. This DAC, it became evident from the number of verticals coming to the fore and the ways in which big EDA is attacking them.