SiFive - September 20, 2016
Why The Time is Right for Open Source Hardware and ‘Chips as a Service’
The open source software movement has been credited as a key driver of the birth of the Internet Age. Without developments such as Linux; the free Apache Web-server platform; and tools such as Java, Perl and Ruby, the Web as we know it would likely not have been possible.
Applying open source principles to hardware to develop new, disruptive systems isn’t a new concept. Garage hobbyists have long been reverse-engineering boards and devices for their home-grown projects. One of the most successful examples can be found in the popular Arduino development board, which helped popularize and power groups like the Maker Movement. The Maker Movement, as defined by Adweek, is
the umbrella term for independent inventors, designers and tinkerers. A convergence of computer hackers and traditional artisans, the niche is established enough to have its own magazine, Make, as well as hands-on Maker Faires that are catnip for DIYers who used to toil in solitude. Makers tap into an American admiration for self-reliance and combine that with open-source learning, contemporary design and powerful personal technology like 3-D printers. The creations, born in cluttered local workshops and bedroom offices, stir the imaginations of consumers numbed by generic, mass-produced, made-in–China merchandise.
It's a very inspiring movement, and I speak from personal experience as you know. In my previous position as the Vice President of Marketing at Atmel Corp, we leveraged the energy and passion of the Makers to fuel growth in the long tail of the company’s business--a tactic that proved to be rather successful (if I may say so). In fact, the community we managed to attract was so large, it resulted in Atmel content to be shared more than the content from 39 semiconductor companies combined. The open source community and philosophy is a great foundation for business, and exactly the reason I joined SiFive as an advisor, and CMO. Let me try and explain that.
"We", in commercial hardware, face a hurdle the software world does not. Where software, once developed, is essentially “free” to distribute, there remains an inherent cost in producing silicon and boards, even when designed by the crowd. The expense and difficulty in producing hardware has historically made this the realm of the largest companies. This has led to a logjam in some corners of the semiconductor world, where only the largest system designers can get the attention of the shrinking number of chipmakers able to produce the custom silicon needed for new devices. As a matter of fact, when I worked for Philips Semiconductors, back in the early 2000’s, building an SoC (system on a chip) required hundreds of millions of dollars, years of continuous work, and a large corporate research institute…now, it can be done by a team of people small enough to split one large pizza, combined with the power of the open source community.
A growing number of companies are looking to change this dynamic through open source silicon. While there are still costs associated with the physical production of the chips, the ability to leverage existing open source assets not only reduces the cost of entry dramatically, but also enables groups previously locked out of the development cycle due to budget, knowledge or location. Open source hardware has the ability to put the crowdfunded project from two engineering students in Nairobi on an equal playing field with a Silicon Valley startup or Skunkworks project at a Fortune 500 company.
The Internet of Things seems like a natural first market for custom silicon: most people agree this category will be responsible for the next wave of innovation and growth in the technology industry, but the diversity of IoT devices could be staggering. In fact, Gartner forecasts that by 2017, 50 percent of innovation in IoT solutions will emerge from startups less than 3 years old. We can try to imagine how this innovation will affect businesses in the coming years, but with new solutions addressing our needs so regularly, realistically we can’t conceive what the IoT will disrupt in the next years.
The functionality and price points of beacons, sensors and other non-consumer connected devices will require low-cost, custom open source silicon.
You might think it is still early days for open source chips. However, the market seems ripe to explore non-traditional options to enable continued innovation. I think, judging from experience with similar dynamics in my past assignments, that this development will go way faster than people can imagine. This morning, EETimes also listed SiFive in the list of 2016's Emerging Companies to Watch. Please watch this space since "Alternatives Really Matter" and SiFive is exactly that alternative. I’d love to receive your thoughts and comments. In the meantime, please join the community at https://forums.sifive.com/