Competition in the electronic design automation (EDA) space is fierce. One minute companies are in a legal battle, the next, one is acquiring the other. Companies in other industries partner with more success – why not EDA? Electronic News sat down to discuss these issues with Ian Macintosh, chairman and president... ... of OCP IP; Phil Bishop, president and CEO of Celoxica; Chris Rowen, president and CEO of Tensilica.; and Hal Barbour, president of CAST. What follows are excerpts of that conversation.
Electronic News: How do you look at your competition? Do you see them as an enemy, potential collaborator, acquisition target or somewhere in between?
Barber: I think it depends on what phase of maturity the market is in. I’ve been through a few phases and in the early phases, it is so much fun because you are creating and selling something that is totally new that no one has – and competitors are more like colleagues. Then, it can mature to a point when it is no longer as much fun because there is a lot more ‘in the trenches’ competition – you know everything about the competition, you know everything about the product, you fight every sale out, but you aren’t selling a concept any more because the concept has been accepted. What’s fun is when you are solving technical problems and doing new things. Your survival depends more on how well you do in solving a customer’s problem and choosing the right things to go after – and not screwing up. That’s a lot more fun than being in the trenches kicking each other.
Rowen: We all understand that having too many players in a particular niche means commoditisation not just because you don’t have any differentiation or leverage in doing deals, it means there is not a whole lot of creativity and differentiation and innovation on behalf of the customer. The other extreme is actually just about as ugly: There’s no market at all – you have a product that is so totally unique that no customer could possibly understand it and buy it. Especially for small companies, the key is to figure out what your trajectory is and where you want to exist between those two extremes of ‘so unique there’s no market’ and ‘so common there’s no market.’ You need to come in with something that is unique, yet ties into the rest of the world, and you need to ride some curve so that you can start small and become big. Fortunately, the electronics industry creates so many drivers for change. Moore’s Law scaling alone is a wave that you can ride from success in a niche to success on a grand scale.
Bishop: I think the difficulty is if you are looking at whether a company is a potential competitor or a partner, the problem is, in EDA to be specific, almost everybody is both. The issue in EDA has been that the channels are so strong – and you need a strong channel to differentiate a very amorphous product – that a lot of times the channels end up having a competition for budget. What could have been a partnership potential if you’d really segmented the market correctly, or really understood what the customer needed fundamentally, ends up being a major competition for budget. Part of the issue in the EDA space is that most companies have always looked as if the glass was half full. They haven’t looked at it from the standpoint of, ‘What value do we provide? What’s the real opportunity here? What’s the total available market of what we can sell into?’ So they’ve tried to line up against someone else and customers have been pretty brilliant at doing that – every kind of licensing model has been generated by buyers and customers smart enough to play the vendors off each other when the vendors should have communicated more. Instead it became wildly competitive. This latest trend where there could be collaboration but there’s actually competition. Someone sues someone else, that’s actually more of an economic reality. That’s someone trying to take someone out, literally, so they can buy them.
Macintosh: I tend to see pretty much all industries as opportunities for collaboration. I think partnerships are bound with the half-empty glass view. We have a very mature and very immature industry. It’s immature in terms of understanding relationships and the opportunities to grow and serve markets; we don’t have that vision in our industry at all. I think we did have some opportunities maybe a decade ago when that sort of vision was around. It’s kind of died now. I think it is true that people get played off against one another and people are largely victims of their customers. I still truly believe today that you can partner with everybody. The truth is that you need to understand your core competencies; you do need to figure out how you pair up with someone. I don’t understand why companies don’t pair up to better themselves. In reality, most guys could combine and work together and grow each other’s business. Let’s face it, a lot of what we do is commoditised.
Electronic News: What do you mean by commoditised?
Rowen: There are two essential elements to commoditisation, or the thing that people imply by commoditisation. One of them is a pricing model – you could just go out on a Web search and you buy the cheapest thing for which the specifications are all completely standard. That we tend to dread because it implies that there are almost no opportunities for added value. The other piece of what is implied by commoditisation is wonderful, which is that the customers understand what they are buying, it’s easy for them to use, that it is user-friendly and convenient. Clearly, we should all be driving in that direction.
Macintosh: The EDA industry has grown enough to be able to commoditise a lot of SOC design work already. You can still knock things out in better ways using some tools than others, but people are able to go at a problem and develop products today using one of several toolsets without necessarily mixing and matching of them, and that has commoditised the design process to a large degree, hence, all the outsourcing. Let’s be blunt about it -- we’ve outsourced the design process to the lowest cost of labour. It’s long gone from the US. It’s gone into parcels around the world where people have pockets of expertise and/or don’t cost a lot.
Electronic News: Does this come back to the industry missing the mark in terms of partnering with the right players?
Macintosh: The thing that has always troubled me in the EDA industry is the lack of standardisation and collaboration. This problem of the design productivity gap we all talk about was caused because of the combative nature of the way people do business. Let’s be blunt, we all know which EDA companies are experts in certain areas, but they all produce these whole products that claim to serve the entire market. Any one design flow on its own, coming from any one major supplier is a pain. It’s not an excellent solution. It could be if you could mix best of breed that you could have an optimum solution and go after this design gap problem. And you have a chance with standards that maybe you could integrate in more readily unique products that have real value.
Bishop: If you go back to the competition versus collaboration aspect, there are two divergent aspects of the EDA space. Some of it is that you’ve got small companies having to innovate, they’re not going to partner with a big guy that thinks he has a commodity flow that, by the way, isn’t very commoditised in the sense that it doesn’t necessarily meet up with any standards or do what customers want. So you’ve got a very strange scenario in that there’s this idea of a certain amount of maturity, but I don’t think it ever really matured very well. It’s never established the business models you would hope to see theoretically, if it had commoditised correctly. You really see the value and the customer doesn’t necessarily see the value, so that sets everybody against each other. I like to look at other companies and ask, ‘Does this company have a set of synergist products to [Celoxica]? Do we call on essentially the same type of customers? Are we looking at the same market segments? And then, frankly, are they big? What’s their culture like?’ Some of those aspects are what lead me to partner with them or not.
Electronic News: Why doesn’t the EDA market do partnerships well?
Macintosh: Here’s an example [picking up an Apple iPod from the table] where people do partner. Whoever branded this and moved it through a channel is not the same guy who designed the electronic components for it, or the same guy who built it - that’s partnering. These people are getting to market faster, they are growing the market share, because they are collaborating – the problems became too big to work alone. Tell me why this is different from the SoC [system-on-chip] design industry. I don’t see how it is.
Bishop: It shouldn’t be. For some reason, the people in SoC design, the people in system design, a lot of the electronic design people in their own segment don’t see the real solution. Each of the areas here [in the iPod] takes a specialisation and intensity that might be somewhat unique. There’s a specialisation that it took for each of the giants to be good at what’s in [an iPod]. For some reason in EDA I feel like a lot of people think that they can just, because the products are pretty amorphous, they can reinvent their software pretty easily to add whatever Magma or Synopsys has.
Electronic News: Does the structure of the EDA industry hold back the electronics industry?
Rowen: I have a lot of respect for the leading companies in EDA, but there is at the root of it, a sense that many people in the EDA industry have lost track of the fact that the problem being solved is getting bigger, faster than the EDA products are growing in breadth and impact. So EDA in a sense is, I fear, represents a shrinking fraction of the solution. I think part of the reason why you see a semiconductor industry that is relatively healthy and an EDA industry that is relatively unhealthy because they represent a shrinking fraction of the R&D dollars being spent. That is not a good sign.
Bishop: Take a look at the external EDA expenditures for Intel and Samsung and Fujitsu. They are actually reducing EDA expenditures and are doing more and more development internally.
Rowen: IP is growing at 20 per cent per year, and EDA is growing at five per cent. What does this mean? It means that the dollars have shifted into IP.
Bishop: The EDA industry is not providing what those customers need. Their market is growing faster, their challenges are growing faster than the EDA industry is able to produce solutions for, and it’s a fundamental problem here. As time goes on, we’ll be seen as more niche oriented.
Electronic News: What can the industry do to combat this?
Bishop: What’s interesting is that you have 50 to 100 startups in the design for manufacturing space, how many of those are partnering with each of the big EDA players? Almost none. Each of them is trying to create its own unique solution. But then you have Fujitsu and Toshiba doing their own tools because they don’t know about the small innovative companies, they only see the channels from Cadence, Synopsys and Mentor, and that’s part of the issue. The big guys in the market as a whole have never figured out how to take the innovations that are coming up and embrace them. Instead, they try to kill it or buy it.
www.cast-inc.com
www.celoxica.com
www.ocpip.org
www.tensilica.com
![]() |