People now understand that while Linux is about Open Source, Open Source isn't just about Linux. Today the Open Source frontier has moved well up the software stack into the application domain. Firefox is picking up market share in the browser space, for example, and MySQL is gaining on proprietary databases.

Like Open Source software, we are now seeing the beginnings of a movement up the hardware stack. Though in its infancy, it's only a matter of time before the Open Source hardware domain experiences a growth spurt similar to its Open Source software counterpart. While we sleep, the adoption of Open Source hardware should open up the playing field and ultimately revolutionize the technology landscape as we know it. This transformation will be most evident in the appliance world of audio and video devices, mobile and VoIP phones, games and entertainment servers, gateways and security devices, robots, PDAs, and the like.

In the past two decades, these products have seen their designs migrate away from their proprietary closed origins into more open flexible formats. Connectivity, interoperability, and security requirements coupled with diminishing product lifecycles and associated pressures to minimize development costs and time have driven this migration. Today the typical hardware design is built around a generic silicon core (x86, ARM, PPC) that's shrouded in a fine Linux/BSD operating gown (although many designers still feel safer wrapping their silicon in a darker embedded Windows cloth. The gown is then embossed with a sprinkling of open software modules (a dash of NetFilter for networking, OpenSSH/OpenSSL for secure access, TACACS+/LDAP for authentication, etc.). Finally a cosmetic layer of highly visible (but extremely thin) intellectual property is applied.

Over the coming years it's inevitable that these hardware designs will open up, and we'll experience a huge wave of Open Source hardware designs. It may take time to push this revolution down to the chip level (given the different capital and timing parameters there), and it may prevail only in the generic hardware market and not in the niche markets - but it will prevail.

To many, the notion of open hardware seems foreign, since we have been conditioned to associate Open Source with software. It can be reasonably argued, however, that open hardware has already had a more significant impact on our industry than GPL, or Linux, or any other modern software. The whole commoditization of the computer industry began with the IBM PC in the early 1980s. It's the force that has driven computing technology into everyone's personal space. A wealth of peripheral devices, plug-in ISA cards, applications, operating systems, services and evolutionary clones were developed that gave the PC a momentum that would otherwise not have been possible. This model still prevails in the PC hardware space where the value of leading suppliers like HP and Dell isn't so much in the IP in their hardware designs, but in the business and support services they offer.

The same forces that are driving Open Source up the software stack will inevitably drive openness down to the silicon core. And the result is that users of the resulting open hardware appliances will benefit from the same compelling value propositions:

1. Standardization - The Open Source model is based on the notion of developing evolving and accepted standards, and this is of greater value in the hardware world. Standards increase interoperability and in today's Internet world, secure interconnection is imperative. Historically the hardware design environment was one bound by regulated and industry standards such as IEEE, ANSII, and FCC. But today's time and cost pressures have reduced the scope for these to be formally negotiated or legislated effectively. There is real opportunity for an Open Source model to take on the standardizing role in hardware design.
2. Cost/value - Another important benefit of the open model is that customers will look to their suppliers to deliver tangible value. Being able to supply quality product becomes a basic imperative to participate in the market, not something you can charge a premium for. So suppliers in the open market are rewarded for delivering associated services and support. Open businesses spend more money on engineering and support and service departments, creating tangible value. They reduce the money spent creating mystery value in their sales and marketing departments. They also reduce the money wasted reinventing the wheel in their R&D units. Standardization brings lower costs.
3. Quality - With an open model, the development process draws input and testing from the widest user community. Peer reviews expose weaknesses that are fixed quickly, rather than just quietly. The results are robust and meet real, rather than perceived needs. Open hardware designs will never offer free solutions, as hardware always involves tangible bill of material (BOM) costs. However, just as is the case with Open Source software, open hardware products will be adopted not because they are free, but because they are better.

There are hurdles, however, that must be leaped before the new open hardware frontier can pick up momentum. One such hurdle is the need for licensing models. The software industry has developed a robust and evolving selection of Open Source licensing models, but they don't meet the needs of hardware.

Also the whole notion of open hardware design won't become credible until a bunch of commercially successful businesses have been built around it. When we can point to wealth being created and sustained enterprises, only then will the open hardware model gain momentum.

In mid-2004, Opengear, Inc., (www.opengear.com) was founded and funded in the belief that open hardware, coupled with Open Source software, is commercially valid. Opengear is investing extensively in okvm (http://okvm.sourceforge.net), an Open Source project that is developing hardware and software solutions for the console and KVM management market.

This market typifies many hardware appliance niches. The market is small (approximately $800 million) with six to eight major suppliers (Avocent, Digi, Cyclades, Raritan, Lantronix, Perle, et al) - all of which offer Linux-based hardware solutions, with very similar features (and at quite similar prices). There is little interoperability and standardization, however. In fact it's quite the opposite, as most of the leading suppliers have developed their own unique RS232 pin-out standards for the RJ45 serial connectors they all use.

On the cost/value front, Opengear's mission is to drive all the hardware and software price barriers out of remote management. The long-term goal of the okvm project is to deliver a way to simply, quickly, affordably, and platform-independently absorb remote management technology into other products. Today you find console and KVM management solutions in major data centers. However, at Opengear we foresee that once the cost of remote management spirals down then it becomes a logical imperative to embed these capabilities not just in server blades and rack infrastructure but also in the smallest of single-board computers, telephone and entertainment devices, even household appliances. Opening up the hardware will also open up millions upon millions of new commercial opportunities.