Experts At The Table: EDA Outside The United States
System-Level Design sat down with Serge Leef, vice president of new ventures and general manager of Mentor Graphics’ system-level engineering division; Vishal Kapoor, vice president of marketing for new business at Cadence; Ron Burns, vice president of sales at EVE; and Walter Ng, vice president of the IP ecosystem at GlobalFoundries. What follows are excerpts of that conversation.
By Ed Sperling
SLD: Our reports from inside China show EDA is almost non-existent inside many companies and that there is rampant piracy of some tools. Will that change?
Ng: China is still growing very fast. We are aware of the EDA software challenges in China as well as the IP challenges. Many IP providers do not deliver GDSII. We have to do mergers in our foundries to facilitate the business of silicon. Companies there are catching up. As they get to more advanced design—and we’re seeing a rapid move in that direction—having the latest tools and capabilities will be more important. Whether the second-tier market will be ‘unofficial’ software or whether they can get the latest and greatest copies becomes more of a question. But we are seeing designs out of China that are 55nm and 40nm. As they move toward the leading edge there will have to be a way forward and having legitimate, up-to-date copies of EDA tools will be more important.
Kapoor: The change that will happen will be as some of the startups achieve critical mass—when they have 30 to 50 people instead of 3 to 5. As they mature and get closer to success they will want legitimate copies. There are two reasons for that. One is just legitimacy. They want to serve not just the local market but the broader global market. The second is support. When you get an illegal copy there is only so much you can do with it. One other aspect we see increasingly in China is they don’t design from scratch. It’s more of an integration focus. As they catch up, they won’t want to stay lagging. They will want the tooling. We definitely see a lot of opportunity for EDA.
Burns: We’re a little bit different in that we’re a supplier, but we do have intellectual property. When we went into China there was a pause because we’re providing transactor models. But today we’re looking at China as a huge opportunity. The challenge is making sure we have the right support infrastructure. We’re already seeing that support requirements are intense. It’s not a question of whether the opportunity will be there but whether the return on investment is there to match that opportunity. But you’re going to see some of the largest chips available at the most advanced geometries coming out in the next 24 months that will rival anything we’ve seen from the United States. We’re talking about processors.
Leef: I have seen this several times where we would go into a major Chinese company to conduct training on a product and 50 to 60 people would show up for a training session even though, according to our records, they have one license. These are some of the major companies, too. We’ve been trying very hard to convince them to go legal. One company in particular in the past year has bought five licenses. It’s not enough, but it’s some progress. We also do a lot with car companies in China. They aspire to go into international markets and we don’t sense any funny business there at all. That’s a different pattern than telecom and semiconductors. We also see a lot more government involvement and nudging forward to create a design infrastructure in China, so we think there may be a fabless design house explosion there.
SLD: How does that compare with India?
Burns: I spent two years at WiPro, so I witnessed this firsthand. India was more support-oriented. Then it became a technology evaluator and driver, which is stage two. Eventually it will become a market of its own, but at this point it’s still at stage two. We have a lot of things going on in India. But India is not yet the driver. They are the technology builder. And in many cases they are the evaluator of technology. As more and more responsibility shifts to India, it will be a very interesting market.
Leef: India has two kinds of companies, either protrusions of multinationals or fairly undercapitalized homegrown companies. With the multinationals, initially India didn’t have much influence but that is changing. There are new chip designs going to India and architectures coming out of India. It is starting to look like a worthwhile investment.
SLD: How about the rest of the world? Does that match what’s going in the U.S., Europe, China and India?
Kapoor: If you add Japan in there the answer is, ‘Not yet.’ If you look at Abu Dhabi, as they prepare for a post-oil economy they’re going to invest a lot of money. They have the patience that even India and China don’t have because they face the classic Western pressures. If something isn’t ramped up in a year it’s probably not worthwhile. But if you look at other areas like Vietnam, Eastern Europe and some other regions, there are people from the Western areas who go back and take advantage of the local region. Will that eventually lead to critical mass that will matter to the rest of the world? Yes, but it will take some time.
Burns: In the Middle East we don’t see a market yet, but we already are feeling the effects of the capitalization of the Middle East. If you talk to VCs you cannot get funded for a chip project here. That’s not the case in the Middle East. There are many ways to get co-capitalization. They are involved in a very big way, like GlobalFoundries, or in a small way with lots of little startups, and in the middle with companies like OpenSilicon. I think we’re going to feel that effect for a long time, but it’s not a market yet. That can change. If you look at Japan, that’s been a tremendous market for us despite the economic conditions. There’s a culture of long-term patient investment, despite the economy there.
Leef: Japan has been very much focused on the back end of the semiconductor process. They’ve invested a lot of effort in verification, place and route and custom ICs. That’s explainable because large Japanese companies all had their own fabs and they were vertically integrated. What we’re seeing now is Japan Inc. gradually moving toward a fabless model, as well. That seems like a precursor to drive more functional design tools. On the front-end they’re basically 1G. Creating advanced methodologies hasn’t been a priority, but as inexpensive access to fabs in Asia becomes more of a reality, a lot of investment and original thought will shift to the front end. For functional verification, the upside is substantial.
Burns: And software is a driver. They are feeling the effects of software content in the devices. For us, imaging and multimedia in all its forms are important. Software is driving this need. We’re feeling that. The size of the designs are not as large, but hardware/software co-verification are creating a range of opportunities.
Ng: We see the EDA companies in certain areas to start some of their design centers. It’s happened in Vietnam. Israel is certainly up and coming. We’ve seen a large investment in Romania and other Eastern European countries. Over time the abilities will grow and there certainly are a lot of entrepreneurs in those regions as well. The Middle East has made a significant investment and will continue to make a significant investment. The road map is for making it a technology center. This is going to be a strategic, prolonged investment. The VC community here has slowed and they look at this as an opportunity. The goal is to develop Abu Dhabi into a technology and manufacturing hotbed.
SLD: For EDA companies, are these pushes into new areas for market seeding?
Kapoor: I don’t know if it’s any different than what’s been done in the past. If there is an emerging region, we want to make sure we have a design center there. You can’t do business in China like you can in other parts of the world. If you don’t understand the relationships and the dynamics that work with some of the government-funded projects, you’re going to be lost. If you don’t capture the understanding of those dynamics and fail to put together a 3-to-5 year or even a 10-year plan, you’ll be in the same situation in several years as when you first begin.
Tags: Cadence, EVE, GlobalFoundries, Mentor Graphics