Everyone has heard of open source software, but can the same principles be applied to hardware?
Some people argue that hardware is so expensive to manufacture and modify that it prevents hobbyists from contributing, and thus stifles the development of an open source hardware community.
This isn’t entirely true. In fact, the huge popularity of community-developed microcontroller platforms (Arduino and its huge collection of add-on modules being the most famous examples) tends to show the opposite. Other examples include the USRP software-defined radio platform, Texas Instrument’s Beagleboard single board computer, or the Openmoko mobile phone (though the latter has enjoyed limited success).
But while those projects feature open and public hardware specifications, “traditional” schematics, printed circuit boards and mechanical designs, the whole semiconductor design and manufacturing process remains a poorly covered area. There are a few pioneers like GRLIB (LEON3), OpenSPARC and OpenRISC. But all suffer from excessive complexity, slowness and large hardware resource usage – if not outright poor or unfinished design. These factors make them difficult to access and stifle their wide adoption, with a need for oversized FPGAs, modern semiconductor processes and advanced logic synthesis tools – all being very expensive.