... or do we need an EDA 2.0 revolution?
Last month, at 47DAC, I participated in a panel that was announced as Lucio’s Litmus Test: Is Your Start-Up Ready for the 21st Century? Lucio Lanza, an industry veteran and Venture Capitalist in the EDA sector started the discussion by asserting that “Investing in EDA is the best way to lose money, slowly but certainly.” Yet, there is a lot of innovation to be done in EDA. There will be a continued need for better and newer software to help engineers build silicon systems in the future.
So, if the demand is still there, what has changed and what makes investment in EDA so problemantic?
Two things have changed over the since the 20th century: the changing culture and the way economics of EDA are changing. First, the culture has changed. In fact: the cultures of both young people and geeks have changed dramatically in the last ten or twenty years. And this is changing all software related industries, including EDA. People have a social life, but an important part of that social life is lived on-line: on blogs, twitter, facebook, youtube. You name it. Young people have very different values from their parents. The generation that has recently hit the markte and that is currently graduating is called Generation Y. GenY-ers are constantly online, texting to friends (and strangers). They live and breathe by the new communication media. Sometimes, this generation is designated as the Digital Natives. Older generation, who have learned how to use computers at a later age, are called Digital Immigrants. GenY has the following prototypical traits:
- job hopping
- tech savvy and eager to learn
- self-assured and independent
The big cultural difference between the GenY and the previous generation (Generation X) has caused major uproar in the HR departments. Some HR consultants focus specifically on how to deal with Generation Y.
To make things even worse (or better), Geeks form an exponent of this generation. Geeks have best friends across the globe. They collaborate on projects with people they have never met, producing results that meet and exceed industrial quality standards. (Think: Linux, Firefox, Wikipedia) Cutting them off from the Internet and their Web 2.0 applications is like cutting a Baby-boomer off from their television and telephone (even though Geeks don’t own a TV or a land-line: they use YouTube, BitTorrent and Skype instead).
Another thing with GenY and with Geeks in particular is that they expect digital content to be free. Generation Y expects television shows, music and software to come free of charge. Geek use the word “free” in an even more radical meaning: “Free as in Freedom, not as in Free Beer.” Geeks want their software not only free of charge. They also want to be able to study, modify and redistribute the source code.
At the DAC panel, people in the audience asked questions like: “If your software if free, how will you make any money? What is your asset?” They are missing the point. Companies like RedHat and Google are making tons of money. RedHat uses Open Source software and Google gives most of its services for free (as in Free Beer). Free and Open Source software are a fact of life. They break some old business models, but they also enable new models. Those who find the best model will prosper the most!
It was interesting to see Lucio talk about these new things, that are changing the world. I have accepted most of these changes while growing up as a GenYgeek, but Lucio is well into his sixties. He has learned to understand and appreciate the world as it was in the second half of last century. He has seen his fair portion of revolutions. And still he is excited about what is happening today and about how the semiconductor and EDA industry can cope with these challenges. It is imperative that the decision makers at EDA companies (and at other companies too) get to know the sub-culture of their new customers. Teens and twenty-somethings have a different set of values than the CMO and CEO of publicly traded multinationals. So if you are over forty and you want to stay in business: talk to your kids!
The second thing that is changing the EDA industry is the way and the size silicon systems are developed. As Gary Smith pointed out on the first day of DAC, the development cost of a single chip is approaching $100M. This cost is so incredibly high that start-ups can no longer afford to build their own chips. VC’s will no longer fund this. The number of ASIC design starts is steadily dropping. I don’t have the exact figures, but it feels like 10.000 EDA sales people are trying to sell million-dollar tools to four or five chip designers. No amount of rock star marketeers is going to expand the EDA industry from 5 billion to 25 billion.
What will happen instead was quite graphically predicted by Gary Smith in the following parable: A man in Africa takes his canoe and paddles to a nearby town. He buys an FPGA from a local store and pays five dollars. When he comes home, he spends his evenings in a shack programming the FPGA so that it can control the irrigation system of his land. After he gets his irrigation system to work, all of his neighbors go buy an FPGA and use his design to irrigate their lands as well.
While this story may not be entirely politically correct, it does illustrate a point. There is an ever growing group of people that are not served well (or at all) by traditional EDA vendors. Small businesses and freelancers have a hard time even getting an evaluation license for most EDA tools. The same goes for small design teams that are part of a large not-so-high-tech company. If you want to sell to this group, you cannot expect your average sale to be over $ 100k. Perhaps $ 1000 would be more realistic! This requires a whole new way of doing business. FPGA vendors are successfully serving small customers along with their large accounts. A new style of EDA company is joining them.
Lucio Lanza deserves credit for his ability to understand this new evolution and his willingness to adapt. I’m sure he will make a good buck in the process. Many others in the field will try to preserve the status-quo. They try to grow and grow and grow, feeding on an ever smaller market. They will hold out for a while, but eventually they will become extinct like dinosaurs.
One way to loose money in EDA is to try and fight the dinosaurs. Another way is to try and join them. The time has long past where VC’s would invest ten million dollar in a start-up, hoping for a 10x exit within five years. The acquisitions we have seen in the last few years were mostly in the order of tens of millions of dollars. Of course, the recent acquisitions of Denali and Virage (for about $315$M each) are spectacular exceptions. But VC’s are still sceptical about the exit opportunities for EDA start-ups.
As Yunshan Zhu from NextOp put it: the most important VCs in a 21st century EDA start-up is the founder’s wife.
After the dinosaurs came the mammals: smaller, smarter and better adapted creatures that eat the eggs of the soon to be extinct former rulers of the world. After traditional EDA comes EDA 2.0. New and existing EDA 2.0 start-ups have the knowledge of social networking, low-touch sales, online marketing. Their size allows them to be lean and agile, to quickly respond to any changing market conditions. This will allow them to provide a nice return on investment for their investors. Including the founder’s spouse.
- A new EDA 2.0 company: Plunify (opinion)
- EDA 2.0 (opinion)
- Can we have an open source simulator? (opinion)
- Why is GHDL (currently) not good enough? (opinion)
- Lacking an open source VHDL simulator (opinion)