Innovation is alive and well in EDA!
We all know that EDA industry growth is fueled primarily by innovating solutions for customers' new, burning issues, and to keep pace, core EDA vendors must continually innovate. Having actively participated in EDA innovation in a systems company, Rockwell International, as well as four innovative EDA startups - Daisy Systems Corporation, Quickturn Systems, Verisity, and now Jasper Design Automation - I have observed that EDA innovation comes in three primary flavors:
- Node/complexity-driven innovation
- Ecosystems innovation that extends existing design paradigms
- Disruptive paradigm innovation that enables movement to a new EDA paradigm, such as gate-level Placing and Routing (P&R), Register Transfer Level (RTL) synthesis, licensable processor cores, and formal property verification
Measured by headcount, most of the innovative energy in EDA focuses on node or complexity-driven innovation. This entails enhancing existing technology categories so they can handle the new physics wrinkles of the next process node or the volume of data and processing loads presented by the quadrupling of available transistors.
This type of innovation is hard work, especially for the big EDA suppliers, since it demands major upgrades of virtually every tool in the tool chain, and all within the two-year cycle of Moore's Law. It's a testament to EDA ingenuity and innovation that the products that were originally developed for circuits of thousands of transistors in a one-micron process are just as effective and accurate for circuits of billions of 20 nm transistors!
Whether the objective is managing the physics of double patterning and FinFETs or building design tool capacity for hundreds of thousands of lines of RTL, the topline objective of this type of innovation is offering users the same familiar methodology in the face of burgeoning complexity.
A fair amount of energy in both big EDA and startup EDA companies is invested in innovation that fills out the ecosystem for an existing design paradigm. This entails inventing and popularizing tools, IP, and sub-flows that make the existing paradigm more efficient. For example, automated Design For Test (DFT), Automatic Test Pattern Generation (ATPG), logical equivalence checking, and static timing analysis are crucial to today's RTL-based methodology. Yet these technologies were commercialized only after HDL simulation and synthesis transformed the design paradigm to the RTL.
Today, we see commercial EDA suppliers participating in ecosystem innovation for verification, as EDA companies add coverage tools, smarter verification IP, and FPGA prototyping solutions. In addition, EDA user companies often author many innovative EDA ecosystem solutions, which are generally targeted at their company's very specific needs.
Disruptive paradigm innovation
The third and most challenging type of EDA innovation is disruptive paradigm change. It's challenging to birth this type of innovation because new paradigms almost always displace existing solutions, discouraging suppliers of existing technology from embracing and advancing the innovation: the classic "Innovator's Dilemma".
A second challenge that goes with disruptive paradigm innovation is timely technology deployment. New paradigms don't succeed unless the critical mass of required features, infrastructure (scripting, IP, and so on), and knowledgeable users can be brought together long enough to realize the ROI of using the new paradigm. Often, the new paradigm offers huge productivity or time-to-market value, but can't be realized without having all the right pieces in place. In my experience working at startups that pioneered hardware emulation, constrained-random simulation, and formal property verification - all disruptive to varying degrees - many iterations of market learning were required to identify all of the ingredients necessary for broad deployment.
EDA Innovation is alive and well!
All three of these types of innovation are needed for a healthy EDA industry that meets its customers' demands. Big EDA companies execute node/complexity-driven innovation and ecosystem innovation on an ongoing basis. Startup EDA companies are also good sources of ecosystem innovation. Disruptive paradigm innovations most naturally fit startup EDA companies, which have no worries about cannibalization of existing product lines. All told, innovation is alive and well in EDA!
 Christensen, Clayton M. "Innovator's Dilemma". HBS Press, Harvard Business School, 1997.