April 24, 2006
Thought Leaders -3 men, 3 minds, 1 industry
Please note that contributed articles, blog entries, and comments posted on EDACafe.com are the views and opinion of the author and do not necessarily represent the views and opinions of the management and staff of Internet Business Systems and its subsidiary web-sites.
To me, the most exciting things right now are in the area of biotechnology, biological materials, and the area of alternative energy. Battery technology and storage mechanisms-these are big materials issues. Biology is also interesting, because there are all sorts of aspects of nanotechnology involved there.
Basically, the physical materials aspects can be exploited in medical applications, or in other biological systems. For example, being able to tag certain types of proteins and then electronically measuring their concentrations in blood samples, injecting the protein tags into humans and scanning for tumors and tumor growth-the medical applications for these technologies already exists, but the materials have to be developed. The materials are still maturing for these end applications.
Q - Turning to EDA, what's the biggest problem facing the industry right now?
Wally Rhines - In EDA, our biggest problem is revenue growth, and we're going to solve it by looking closely at two problems. Because if you look at where the growth is occurring in the industry-and even these areas are still in their infancy-it's DFM and ESL. Both of these are very small today, but are growing rapidly. And in both cases, there's a clear reason why these two areas will be more and more necessary going forward.
I personally find DFM to be the more exciting of the two-that's where things are happening, and people are waiting around and watching to see what others in the industry are doing. There's real growth and real commitments being made in the DFM space for the reason that the dollars at stake are so large. And I believe that's far more the case than for the opportunities at the ESL end of things.
Today, this idea of manufacturing people having to get involved in how design is done is no longer just an idealist's wish for the future. DFM processes are getting implemented today, and the more aggressive the company, the bigger the implementation.
Q - A number of people have suggested that there's a point in the near future when one or more of the foundries will buy up one or more of the big players in EDA in order to consolidate towards an IDM business model?
Wally Rhines - IDMs formed out of foundries? It won't happen! What makes the foundries so valuable is their ability to deal with a wide variety of customers. If you only had to deal with one customer, you could be an IDM. You could have a relatively restricted set of flows, tools, processes, and you could, maybe-if your company was big enough-efficiently load your fab.
But foundries derive their business from a wide variety of customers, so they can maintain the facility at capacity. They are anxious to support a wide variety of design styles and applications by focusing on a limited number of processes. By satisfying a fairly wide variety of applications and design flows, they are able to serve a big enough market that they can be very cost effective compared to an IDM.
Certainly, for the leading foundries there is a certain degree of nudging towards standardization. Their customers would like them to support a wide variety of flows, but the foundries would like to minimize that breadth. So, of course, the foundries try to convince their customers to use a smaller number of variabilities.
But efficiency in manufacturing is very important-it's quite costly to support different manufacturing processes. So how the design is done is not the question, that the foundry has to support models for various diverse tools is not the issue. Because at the end of the day, a GDSII file is a GDSII file regardless of where it comes from. For the foundries, the cost of supporting diverse design flows is less by an order of magnitude than the cost of supporting additional process diversity.
Now there have been movements in the past to do what you're talking about-for the foundries to become a supplier, more like an ASIC supplier. The foundry takes a design during the design process and performs design services, or provides a set of approved design tools and gets the customer to use this flows rather than a flow they developed on their own. But in those previous attempts, the foundries actually gravitated back towards being foundries.
There was a time at TSMC when the company had built up quite a design group. But newer management came in and successfully pushed TSMC out of that business and back to being a foundry. Similarly at UMC, they spun off their design activities. Faraday was seen by some UMC customers to be in conflict, and they didn't like that the foundry could create designs that would compete with theirs.
So if not services, why shouldn't foundries be in the business of providing pre-approved EDA tools? Well, some would go along with that, but it's not easy to support EDA tools. How are you going to support everything that people are using today? Customers have shown that they want a wide variety of design tools. Which ones are your going to support as a foundry if you have only provide one of each kind of tool to your customer. What about the other 399 tools they would like to use? What are you going to tell the customers about those other tools in their flow?
It would be a great difficulty for a foundry to manage. If a major foundry narrow itself into a set of design tools that only covers perhaps only 20 or 30 percent of the market, it quickly leaves 70 to 80 percent of the EDA software to be provided by someone else. But the foundry still has to support those other tools if they want those customers, so what would they have accomplished? It takes about 9 nanoseconds for the situation to disintegrate completely.
It may look nice from the outside to be in the EDA business, but every industry is nice from the outside. It's always tougher on the inside, and unless you're going to change something fundamental about the industry, it's always going to be hard to become best in class. The foundries know that.
Q - Changing topics again-what do we do about China?
Wally Rhines - We don't 'do' anything about China. It's simply a great opportunity!
And the biggest part of that is the multinationals that are doing work there. For the EDA vendor, it's a big task to support all those remote sites that the multinationals have going on in China, and to be supporting India at the same time, as well. Although, certainly as far as manufacturing is concerned, there's much more going on the China than in India.
But the bottom line is that just looking at the multinationals alone doing business in China-without even taking into consideration the regional Asian companies and the Chinese companies themselves at work there-China's a big part of what's going to be happening for the EDA industry going forward.
So, what do we do about China? We are going to take advantage of the great opportunities there, starting with the multinationals and moving on from there. China is one of the fastest areas of growth for us here at Mentor, and although the market in China is still small in absolute numbers, it's currently in the $10 million to $20 million range for us-it's not inconsequential.
Q - What about software piracy in China?
Wally Rhines - This is not new. We've had to deal with piracy in the past as design activities have spread west. For instance, it used to be that piracy was a big problem in Taiwan. But as those companies started to go public, they started paying for the software-it was a natural process. There probably still is piracy in Taiwan today, but Taiwan is also a big paying customer in EDA.
I fully expect the same thing to happen in China. As the companies there start to go public, they'll want to publish certifications that say they're paying for their software. It's a natural part of the process. Having said that, piracy is still a tough issue. We really can't wait until everyone goes public for software piracy to stop. But the motivations for using legal software are growing and the EDA companies themselves are making it more difficult to pirate the software through features built into the code.
You can find the full EDACafe event calendar here.
To read more news, click here.
-- Peggy Aycinena, EDACafe.com Contributing Editor.
Be the first to review this article