SMALL FORM FACTOR FORUM
Help Wanted! Industry Leadership
Page 1 of 1
The untimely death of Steve Jobs should serve as a reminder to us all that the individuals whose energy, creativity and thirst for innovation brought about today’s Small Form Factor industry are all getting older and either approaching or have already passed retirement age. The folks who brought us the pervasive embedded computer board technologies of the past 30 years or so, VME, PC/104, cPCI and others, are getting ready to hang up their slide rules (or in some cases already have). And just as Apple struggles to try and replace their irreplaceable leader with a passionate and innovative team, the suppliers and standards organizations that together hold responsibility for the evolution of the SFF industry need to identify and put in place a team that can define and drive the next generation of SFF standards.
Unfortunately, as we look across our industry, we just don’t see it. The last truly innovative step in the SFF community was the introduction of the Computer-on-Module (COM) concept some 11 years ago. As the SFF community has evolved since then and wrestled with the introduction of new bus technologies forced upon us by chip vendors with little system OEM interest, no individual or group of individuals has stood up to pull the community together. Instead, we have one organization that never met a pin definition (type) they didn’t like. We have another organization where suppliers battle mercilessly to get their proprietary technology adopted by their competitors as long as they have a huge running start to give them a competitive advantage. And an entire geographic region with suppliers who disdain standards completely, declining participation in industry trade groups while they continue to churn out one-off customer project-based proprietary solutions or knockoffs of yesterday’s technologies.
This lack of leadership is reaching a stage where it will begin to put a crimp in the future growth of the SFF marketplace. Just as Apple moved their Mac family from Motorola’s “Power” architecture (PowerPC) to Intel architecture processors some years ago, the SFF market is largely based today on Intel architecture processors. And just as Apple disdained Intel architecture in favor of the low power consumption of a custom RISC processor for their mobile i-products, so the SFF market is starting to see the introduction of more and more non-Intel architecture embedded boards. All in the complete absence of a standardization effort of any kind.
It doesn’t take much foresight to see that movement of SFF products to processor architectures other than Intel is probably the next great pervasive technology for the SFF market. Intel just hasn’t been able to get where they need to be with respect to cost and power consumption for the smallest form factors. Power and heat are among the biggest challenges facing the broad swath of embedded OEMs today. However, as RISC processor architectures emerge for SFF boards, we seem to be faced with one proprietary solution after another. Interoperability—Oh Please! Ecosystem? Eco what?
Some will argue that these are not important issues and that we don’t need a common interface, with a standardized bus architecture and form factor to support off-the-shelf I/O. Build a baseboard with your own I/O. Change it as necessary to reflect product lifecycle changes. Tell that to the medical and aerospace OEMs with huge certification costs and timeframes.
Somebody or some group has to stand up and drive this issue; some set of suppliers (and OEMs) who are willing to set aside proprietary advantage to do what is right to grow the SFF market for everybody. Mind you, this is not an easy pill for egocentric board manufacturers to swallow. It’s natural to want to hang I/O for non-Intel processors on the processor local bus, and every RISC SoC is different. Some kind of chip-based core is probably required to bridge the gap. This is a solution that some group will have to develop and “contribute” to the industry, royalty free.
As we look out over the horizon, we don’t see any individuals or groups willing (or with enough clout) to take on this challenge. We see a bunch of folks fighting to use standards groups for their own competitive advantage, and a few go-it-aloners hidden behind URLs and logo clubs. Unless this changes, when the next generation of embedded “gurus” approach retirement age, there’s going to be a whole lot of discussion of missed opportunities and how downright awful it is to do an embedded design.