Clearly there’s EDA software use in India and China, but is there EDA software development going on in those locales as well?

Who better to answer that question than Dataquest EDA Analysts Nancy Wu and Daya Nadamuni, and I was fortunate to have some time to talk with the two of them before we all left the Bay Area and headed south to DAC 2004 in San Diego. Prior to our conversation, I received a brief e-mail response from Nadamuni with an answer to my question. The response was basically:

There’s no EDA development of note going on in either of those geographies, but there is a great deal of third-party tool use going on. If I still wanted to have the conversation, we could talk about that.

Although I was disappointed, I was still interested in addressing the geographies. So for our conference call on June 2nd, I asked instead about the level of design innovation going on in India and China. We started with:

Nadamuni: "We don’t believe innovation happens only in North America, but takes place all over the world. In the last 25 or 30 years, research in North America and Europe has been at the cutting edge of technology. However, both China and India are putting out large populations of new engineers every year. [Of course], some of the engineering colleges in these countries are better than others - and you would expect there to be a correlation between research in those universities and the quality and quantity of the work.’

"A lot of the focus [in those colleges lately] has been on bringing engineering education - both the research and the course content - up to par with the top-level teaching universities in North American and Europe. Right now, however, the kinds of engineers coming out of the schools in India and China are either PCB designers or are IC designers working at 0.18 micron processes or above."

Wu: "[It’s important to note that even today] a large majority of designs in North America are not necessarily at or below 0.18 micron. Some percentage of the design engineering teams claim to be working at 0.18 or .013 micron, but they’re not actually putting many of those designs down on a die."

Q: What is the status of the university pipelines producing the engineers to meet the demand in China?

Wu: "As far I know, there are a lot of good schools in China. After engineers graduate from those schools, they’re solid in the training and theory, but they’re missing the hands-on experience which comes from having local companies who collaborate with schools to help students develop the skills they need before they graduate."

"So far, from what I’ve heard - even though a lot of things have been improving in China, they are trying to enlarge their engineering school faculties and student bodies – a lot things just can’t be done quickly. You have to give them time to improve from one point to another."

"Just trying to get educational programs and collaboration [up and running] between the schools and the local design companies [is complex]. It requires funding and the benefits of [that effort] won’t be in place for at least another 3 to 5 years. We have seen a lot of press releases about engineering students graduating in China, but it doesn’t mean they have the management skills or the funds [to be effective yet]."

"A lot of the start-ups you see today in China are formed by management teams with 10 or 20 years of experience in Western companies. They may bring the management skills and the funds to hire a bunch of local design engineers with less than 2 or 3 years’ experience, then they form a design company and try to find a market for themselves. [Of course], you can hire a bunch of engineers with 2 or 3 years of experience, but it doesn’t mean they all have good skills or the quality of experience to work on complicated or higher-end designs - especially compared with the skill of students graduating in Western countries."

"So what I’m seeing in China - although there are a lot of design companies there, the big ones are not doing so many high-end designs, while it’s the small ones are trying to stretch themselves. They’re [struggling to establish] their marketing and sales channels, and everything else that they need to figure out to keep their business sustained. The large companies are surviving, but the small ones are trying to do the more high-end design, but have problems with how long they can stay in the market."

"Even though a lot of people in China are looking to do higher-end design, those designer are not being produced in the wafer foundries in China. However, we do see engineering teams starting to do more challenging designs and larger designers. They’re [definitely] moving up from PCB work to ASIC work."

"[Ultimately] however, you have to look at the real world instead of just the numbers. There are not so many real products coming out of their fabs and they’re falling into more [and more of a situation where they have to import the products]."

Wu: "What we’re hearing from local engineering and design services is that they’re looking at customers in China who want the fastest selling products in their local market. The demand is so high, the Chinese just can’t manufacture fast enough to meet the need from inside the country. Therefore, a lot of these chips are being imported from Japan, Europe, and the U.S."

"So, it’s not necessarily an advantage in the China market [to be manufactured on the mainland]. Right now there are products coming in from all over the world - from Japan, Taiwan, North America, Europe - all aiming at the PRC market. The competition coming from outside Asia is high enough that even though there are local design start-ups in China, they have to find a way out to maintain [their competitive edge against foreign competition."

"We believe the majority of the products being manufactured in the PRC are at 0.35 to 0.5 micron. However, there’s no real concrete data on what they’re actually doing. SMIC, for instance, is claiming they’re moving into 0.18 and 0.13 micron. You’ll see that they’re building another fab in Shanghai, but I believe it’s still under construction. I don’t think they’re at 0.18 or 0.13 yet. And, they’re still importing designs and not depending on the local design talent."

Nadamuni: "Most of the expertise in India is in software. TI, in the late 1980’s for instance, set up shop in Bangalore, but they were mostly doing software. Only in the last 5 to 6 years have you seen some hardware design going to India [to any measurable extent]. Now there are at least 25 companies from outside of India who are doing hardware design in Bangalore. There’s also work going on in Hyderabad, etc., but the main focus continues to be in Bangalore."

"And, now it’s not just off-shore design services. If you look at HP, IBM, Intel, Broadcom, Conexant - you name the company, any semiconductor or systems company in Western Europe or North America - they all have design centers in India. Many have only arrived in the last 3 to 5 years."

"However, when you’re talk about the term ‘outsourcing,’ you’re really talking about handing a design off to a third-party or having access to the physical body to assist you in the project. However, when you’re a Motorola who has Motorola employees in India doing design, you can’t really call that outsourcing anymore. You’re really talking about ‘in-sourcing’"

"Motorola is doing hardware and software design in India. Is that ‘off-shoring?’ Is that ‘out-sourcing?’ No, it’s not ‘out-sourcing’ if it’s Motorola in India – Motorola employees - that are doing the work. It’s ‘off-shoring,’ but not ‘out-sourcing.’ Whether it’s more or less competitive [to do things that way], is a decision that Motorola has to make."

"Historically, companies have sent their less-innovative, or second-generation disjoin projects off shore. But even that is changing right now. Think about it. Look at IBM Haifa. Look at Verisity coming out of Israel. Is that ‘out-sourcing?’ Is that ‘off-shoring?’ I don’t really know. Who does?"

Nadamuni: "Most of these non-international start-ups [you’re seeing in India] are people who have gone to the West, graduated, worked there, have learned best practices, and have taken them back to India to try to make a start up. They’re trying to train local engineers in those methodologies. Can they take the time and the money to train those engineers? Well, it depends on the skill sets that are available."

Nadamuni: "Right now we don’t have enough funding available for innovation here in the U.S. [If you examine things, you’ll see that] most innovative technology development has been funded by corporations and governments. That’s what’s given us the edge here in the U.S. What will kill us here is the lack of funding."

"You can only hope that the other governments in the world don’t ramp up their spending to fund research and innovation. Eventually, however, you’ll see that whether it’s Bangalore or Shanghai, there will be no cost benefit to out-sourcing design. [Even now], people are starting to turn to places like Bulgaria, Algeria, or Tunisia to do design because the labor in India is quickly becoming too expensive. In the end, however, you always have to go where the skill set is, and not where the labor is cheapest."

Wu: "Lots of countries want to take advantage of cheap labor when they can have access to it. However, when engineers’ salaries go up, the companies will move on to other places to find even cheaper labor. It’s a game that you, as a vendor, have to choose to sustain. You have to decide if you’re more interested in the short-term or the long-term gain. That’s the trade off between having the money stay here in this market in North America, versus [attempting to move your customer base] off shore."

Nadamuni: "Cadence does have some R&D work going on in India. Sequence and Magma do as well. Synopsys has a strong sales force at work in India."

Wu: Most of the three big guys in EDA have training centers in the large cities in China. They’re sharing their technology with the universities to train students to use their EDA tools. For the non top-three companies, however, most of those companies are still only in the process of developing sales channels in China."

Nadamuni: "We know that just about every 10 years, an inflection point in the technology comes along. The old guard in an industry undergoes change. We saw [the end of] the Daisy-Mentor-Valid era, and now we’re [facing] the next historical hump."

"[Meanwhile], there will be a sharp spike in design capability in China, and we’ll come to a point where we will need to provide a different level of [sophistication in] the tools to that market. Of course, there’s never going to be a revolution in tools that are being used - it’s always going to be an evolution. The current tool vendors have a ready market in China, but that will gradually evolve to something more demanding."

[Editor’s Note: Please stay tuned for Part II of this discussion.]

*************

June 24, 2004

Peggy Aycinena owns and operates EDA Confidential. She can be reached at peggy@aycinena.com

Copyright (c) 2004, Peggy Aycinena. All rights reserved.