EDA Resurgence Through Open Flows

The EDA industry can experience a resurgence if truly open interfaces and open flows are embraced.

popularity

A couple of years ago I would talk to investors about startups in the systems space and overwhelmingly they were extremely negative: startups were dying; there was no investment; it was difficult to get products to market. This same conversation has happened in the past with infrastructure software, and is happening now with EDA in terms of stagnation, lack of innovation and lack of investment. Rather than dwell on the problems however, I want to see what can be learned from a resurgent systems space.

The revival of the system startup
A good place to start our exploration is to understand how the system startup went from being something on deaths door, to possibly the most vibrant area of the startup scene in just a short couple of years.

A major part of the revival has been on the funding side. While traditional investors had given up on systems, it turned out that cutting edge consumers still loved the idea of cool new things. Thanks to the likes of Kickstarter, we saw many new systems go from concept to overnight successes. This provided funding and much needed visibility, and several of these have gone on to secure larger rounds of traditional funding or obtain large exits; Pebble and Oculus Rift being two of note.

More importantly however has been the role of open hardware and open software in enabling these businesses.

For years we have seen the open software movement revolutionize software and data center infrastructure. Taking an industry that was stagnant and propelling it to much greater heights, with companies building on each other’s efforts, and focusing on their unique innovation.

As they say, ‘A rising tide lifts all boats,’ and in this case as the pool of common infrastructure rose, so did the fortunes of companies building on it.

Open hardware shows the way for systems
A common mantra for people during the software revival was that this approach would not work for Systems, Semi, and EDA. Our products were too complex, too expensive, and required a unique, closed approach to technology and business.

It does seem that open hardware has shown this to not be the case. What began with systems companies adopting Linux, then Android, has expanded to companies building whole ecosystems on open software stacks as well as open hardware designs. These are designs that people are free to evolve as needed, and the companies that have embraced this trend have been able to take a product to market for much lower cost, much faster, and yet still innovate and be profitable.

I am not suggesting that companies open their innovation. Just by participating in this open movement they expand their opportunities. Apple, Adobe, Intel and others all leverage open source in their products, enhance those open parts, and then build closed innovation on top.

How this relates to EDA
Let’s be honest, EDA is broken. Over the last couple of decades I have witnessed it get ever more closed. As this “closed” approach has increased, so has the stagnation. New startups in EDA are forced to beg for access to closed technology which is rarely given, and more often than not they reinvent the wheel and run the risk of legal action. I challenge people to name one other industry where a startup would be required to show a 10x benefit to be taken seriously. Yet in EDA the complexity of integrating with closed technology leads to just that.

Open EDA, grow the market for all
I am not suggesting open sourcing tools – although the open hardware movement has helped revive long dead EDA open source – just opening interfaces and flows. It’s time to learn from the software and systems space revivals that building on each other’s work is the way to expand the market for all.

A good place to start is with truly open interfaces and open flows. Not the, “Let’s put something in a committee to delay the industry for several years so we can milk our technology” type of open, but rather dynamic, open collaborations. This will help establish a common base that all can work from. From this base we will see, much as with systems and software, companies donating useful but non-innovative technologies. Startups can then easily integrate the parts and be integrated into a flow. Now a 2x or 3x innovation can be easily adopted and everyone wins.

Probably a harder part to change will be the business models. Today the big EDA companies try to lock out as much of the budget as possible, abstracting away what technology the customer is actually paying for. This “all you can eat” approach makes it difficult. However, if we can reduce the complexity of integration, quickly show real value, then unlocking new money becomes a significantly easier task.

In full disclosure, I have a natural bias to this approach, and currently work with a number of companies who are building their businesses on open interfaces and leveraging open software solutions. These companies and their customers see a lot of value in this approach but these same companies are stifled when interfacing to the rest of the flow — wasting internal and customer resources begging for access, hacking together fragile solutions, or reinventing the wheel. Startups can only go so far on their own; they need your help.

Taking the first steps
How do we get started? The only place to start is for users to force the EDA vendors to open up. The big three will only respond to customer pressure and customer dollars. As the users of the tools you need to make this happen, and you will ultimately benefit. A small amount of effort today will pay off at a much higher multiple in the future. The vendors need you as much as you need them – so don’t be afraid to apply pressure.

Let’s use your success in driving a resurgence of systems to drive a resurgence of EDA.


Tags:

  • Eda guy

    We need more than open, we need common, I don’t need another upf/cpf, POCV/socv where I have multiple standards.

  • Neil Hand

    I fully agree.

    That was my point with regards to the “truly open” as in everyone working for one. The whole CPF/UPF like so many other initiatives in EDA was about locking down technology under the guise of standards. Keep everyone busy for a few years while pretending to care.