Our industry is going through a big change. The design complexity forces a paradigm shift from a collection of point tools to providing large, integrated productivity solutions. The mechanism is simple yet brutal: SoC system complexity is increasing with Moore’s Law, while at the same design costs continue to rise steeply. The focus in EDA needs to be to relentlessly reduce that design cost. The good news is that we have addressed this challenge and that we can provide our customers real productivity improvements. Let me add that it’s not enough to just have ‘holistic tool integration. If one point tool is in the flow is bad, you cannot be stronger than the weakest link in any placement, routing, analog-mixed signal design system.  This means that an EDA vendor simply must be best in class for all its tools to survive in this market. In short: anything you make has to be significantly better than what the competition is making. This is especially true in the very competitive landscape of EDA. If you’re an EDA company or an EDA start-up and want to succeed, then you have to differentiate. That differentiation has to be seriously better than what your competition.

If you look at look at the enabling technologies, we need to go back the relentless driving force of Moore’s law. We simply cannot understate the tremendous size and complexity of the designs at our customers.  I think that EDA has no choice but to exploit hardware parallelism even further.  EDA as an industry has claimed to do some parallelism. The level of parallelism has been limited to just 4 or 8 processors. In practice that yields a dismal 2.5x overall improvement in productivity; perhaps 3 if you’re lucky.  We have to address that. In analog, for example, Magma has a tool called Finesim, that with 30 processors will give you a real 30x improvement instead of saturating in the low teens. We need breakthroughs such as this across the board. We see similar challenges in verification, which – to be frank - has to get its computational act together to get throughput advantage. I’m aware of a lot of technology in that area. In some cases Magma is actively contributing, in some cases like logic simulation, it is not in our focus and in some cases Magma has played the role of an investor in some of those technologies. In EDA we need to go back to the basics of improving productivity significantly.  We talked a lot; we now need to walk the talk.

Some of the sex appeal of our industry is a little bit in jeopardy. Part of that comes from the fact that it’s currently much easier to do an Internet start-up than EDA. When I was graduating, at that time the coolest thing in the world was to be in VLSI.  The best kids in my class fought for the openings in any semiconductor company. It’s a challenge to bring back that that oomph that we had in 1990ies. We need to rebuild that as a community and attract young talent back from the ‘Googles’ of this world. The algorithms that we apply in our technology are much more complex and interesting than in any other branch of the software industry. 

Well, I just asked my 13 and 14 year old daughters: “do you want to become an engineer?”. They think that biotech is cool, perhaps bioengineering, or do all kinds of small iPhone apps, or even do fashion design. But hard-core engineering is not high on their list. There is an image issue there. I think we need to reach out to make them understand what our exiting industry is about: that we contribute a very important piece of to the dreams that people have. I have tried this with my kids. To be frank, it’s a very steep curve right now.  And to climb that curve, we as EDA industry need to collectively improve our image.

That’s right, we are the central component in the supply chain for green technology, in any form of solar applications to smart power grids.  And I think EDA needs to branch out a little bit. We need to take a bit of risk in doing that.  I mean, we wrote a tool for doing yield management for solar fab applications that is being deployed at a lot of customers right now.

The challenge is not just marketing. In fact, I think marketing is probably just one of the smaller challenges we face. We need to step up as an industry. We have been very dysfunctional. We fight each other so hard that it devalues - pun intended - the great products that we deliver. It’s a race to the bottom that we must elevate ourselves out of. 

This is a tough business in which we have to fight hard to compete. At the same time we have to make clear that we provide an unprecedented level of automation and value to our customers. I think we have to step up right now because they need to understand that quality comes at a price. Top notch engineering talent is needed to build the EDA tools that the information age needs. Without that next generation of chips will not be effective.  It’s important for the industry as a whole that EDA remains successful.

Absolutely, we simply need the very best. We need to bring in fresh outside new talent, with energy, into the business. That’s where we need to think out of the box and execute on to ensure a thriving EDA industry.