By Dean ~ November 6th, 2008. Filed under: Industry, Observations.
Paul McLellan, the former CEO at Envis (coincidentally, his replacement was just announced last week), wrote a provocative opinion piece for EETimes. Mr. McLellan asserts, as have many others, that the current model for the big EDA firms is simply broken. At best, it’s outdated. At worst, it might not be fixable.
One of the most interesting points in his article comes when he describes EDA’s role in the global electronics industry.
“The world’s $3 trillion dollar electronics industry is largely dependent on the world’s $400 billion dollar semiconductor industry which, in turn, is dependent on EDA, a tiny five billion dollar industry. EDA technology is a requirement for this value chain. But the EDA industry, as currently structured, might not continue to be the place that continues to provide it.”
He also shared another very compelling way of putting this all into perspective:
“… the cost of one of its [TSMC’s] fabs is more than the market cap of the entire EDA industry.”
Boy, that certainly gives EDA a sense of its place in the world.
Mr. McLellan continues to catalog the challenges faced by the industry:
- Their customers beat them bloody on price.
- They’ve cut back too deeply to maintain an in-house innovation culture.
- With fewer chip starts, their customers are either consolidating, or going away. In either case, there aren’t as many customers.
- For many product categories, the real value is added in system integration and software, not the underlying semiconductors.
His scenarios for how the industry might evolve are what really caught my attention. I was particularly drawn to the second possibility:
“EDA can become part of a larger industry, say, mechanical design or enterprise software.”
Having grown up in ICCAD/EDA, I guess that I just assumed that we were the center of the universe. What’s become clear over the past several years, however, is that electronic design is a component of overall system design. As the focus shifts to complex higher level systems, and the required levels of abstraction move even higher, traditional EDA tools will still play a role. However, the next wave of “must-have” tools will take a whole system perspective and will need to allow designers to quickly make tradeoffs among different implementations. This is where true Model Driven Design will find its calling. Maybe the big players will figure this out and get serious about the space. My old colleagues at Mentor Graphics have certainly started saying the right things. On the other hand, maybe the “big 4” will keep focusing on increasingly niche tools while somebody else takes a leadership role.