Eric Weddington, an Open Source Architect at Trimble, recently published an article on LinkedIn Pulse that examines the possibility of making Integrated Circuits (ICs) and microprocessors open source.
“What would it take to do that? Imagine it would have to take small groups or individuals, perhaps University teams, to get their hands on some semiconductor manufacturing equipment,” he opined. “Perhaps old equipment sold (relatively) cheaply. And methods used to manufacture chips opened up and shared with a community.”
Image Credit: Derrick Coetzee (via Wikipedia)
Improvements to the equipment could be made, says Weddington, with those optimizations ultimately released as open source hardware. However, as Weddington emphasizes, the industry does not (currently) offer any real incentives to make the semiconductor manufacturing equipment cheaper or easier to use.
“Software tools to design chips would [also] have to be created and made open source so the community can use these tools in conjunction with the re-purposed manufacturing equipment,” he continued. “Design libraries would have to be created and open sourced to pull together parts of a chip design for an IC.”
Despite the obstacles, Weddington says integrated circuits could be the “final frontier” of open source.
“Because if it is not, then you can make PCBs on your desktop all day long, but real control of your electronics will always be in the hands of the few,” he added.
Perhaps not surprisingly, Thibault Cantegrel, Director, Developer Program at Sierra Wireless, believes open source hardware could play a major role in helping the Internet of Things (IoT) evolve.
“The Internet of Things has a pressing need to reduce the daunting complexity of building every piece of an IoT solution — the hardware, the embedded software, communications layer, server and associated software,” he explained in an article posted on M2MNow. “This requires a variety of engineering skills. And open source hardware, with such big communities, well designed and well tested hardware pieces would be very compelling if they fit [certain] requirements.”
Indeed, as Michael Cooney of Network World recently pointed out, open source hardware is now roughly in the same place as open source software during the mid-1990s.
“What made open source software acceptable for many businesses was the arrival of support for it, such as Red Hat; something similar may take place with the hardware.”
Steven Woo, VP of Solutions Marketing at Rambus, expressed similar sentiments.
“A number of major industry players would likely be willing to consider backing a new semiconductor paradigm, particularly when it comes to the IoT,” he said. “As Cantegrel notes, it would have to meet certain requirements, and the monetization aspects of such a paradigm would have to be worked out and clearly articulated.”
Although launching and successfully executing such an initiative wouldn’t be easy, Woo emphasized the idea shouldn’t be dismissed out of hand because of the difficult challenges the industry faces in meeting the promise of the Internet of Things.
“We’ve seen just how much the semiconductor model has changed over the past decade, especially in the mobile space. The semiconductor industry has done a good job of adapting to meet the needs of new markets and applications, and the Internet of Things will be no different.”
When asked to identify a potential catalyst that could help spur the adoption of an open source semiconductor model, Woo pointed to the lack of adequate security in the silicon space and a growing interest in effective provisioning.
“With the right level of backing and contributing expertise, RISC-V could provide a foundation for exploring the open source semiconductor model,” he concluded. “A RISC-V implementation augmented with security features could allow broad adoption and deployment for a broad range of IoT applications and infrastructure.”
Interested in learning more about RISC-V? You can check out the project’s official page here.