SMALL FORM FACTOR FORUM
Page 1 of 1
Across the world, disgruntled citizens have gathered in encampments to protest joblessness resulting from rampant corporate greed and the bloated government that continues to fuel it with quid pro quo. The movement began with “Occupy Wall Street” and quickly spread to other cities. Bailouts for bankers with ballooning bonuses (the ultra-wealthy 1% of the population) are essentially paid for by the other 99 percent—the taxpayers.
Meanwhile, our embedded community has its own “golden rule”—those with the gold make the rules. In this stretched economy, most suppliers and system OEMs alike are starved for resources, barely getting by on the backs of employees who wear three hats to make ends meet. The direction of small form factor (SFF) boards rests on the shoulders of those few vendors who are willing to invest their time and resources in the future of the industry.
Once rich with diverse membership and widespread participation, trade groups have been recently dominated by member self-interest and individual egos. Rather than commissions and bonuses, the rewards are originator fame and unfair product head starts in the market. Many of the market-savvy old guard members who understand from where the industry came and the importance of legacy interfaces and embedded market firmware have long since moved on. Now most of the big name consortia are driven by just a handful of chipset-oriented engineers. These Lehman Brothers- and Bear Stearns-style engineers are developing standards that cater to the highest-performance 1 percent of applications. Embedded ecosystems are being forgone in favor of proprietary self-interest. The other 99 percent of embedded apps no longer have a voice in the standardization process.
As a user of embedded technology, you are not required to consume passively whatever these groups dish out. Luckily there is still time to make changes for the better and restore the feedback loop—surveys, market research and customer involvement—that once regulated these consortia. Just as banking deregulation has taken decades to evolve into the current mess, the embedded market is only partway into the latest technology iteration. Embedded technology standards go through multiple stages between the formation of a new working group to wide market acceptance of the technology. Many new board-level standards from various consortia are still in their infancy. Still others, developed outside of any formal trade group and without any multi-vendor review, might never result in products from any company other than the originator.
COM Modules first lost touch with the broad application base when ETX transitioned to COM Express. Low-power +5V modules with legacy support gave way to multi-core, legacy-free PCI Express modules with enough +12V pins to support 188W of power consumption. The thermal solutions are an exercise for the user, of course. Ironically, nearly every module ships with a BIOS that initializes one of the many legacy super I/O chips that customers almost always add to the carrier board. Better pick the right one! Some of the power pins have been “reclaimed” in the Revision 2.0 specification, which means in lay terms that the power consumption has been traded for the latest chipset features including three digital displays and more PCIe lanes for the “one percent-ers” while foisting an incompatible pinout type upon everyone else. The other 99% of real-world apps are stuck with the carrier board re-design bill without any net feature gain.
Not to be outdone, the governance in charge of stackables also appears to be obsessed with power-hungry dual-core and quad-core processors and I/O, trying to compete against expensive VME/VPX systems rather than keeping the affordable 386/486/Pentium level of performance for simple control apps and ecosystem compatibility. If you’d rather upgrade your military and transportation systems to dual-core Atom at 15W than to quad-core Core i7 at 45W, you might want to tell someone now.
Legacy system upgrades carry a lot of clout with vendors and can be used to bargain for desired features with some attention to developing a simple focused message (sound familiar?). But it has to start with the recognition of the problem and willingness to get involved to nip vendor ambivalence in the bud. There is no magic antidote. It will take a lot of hard work and involvement by many system OEMs. The voices of “we the people” need to be heard in the form of our own embedded “occupy” movement. “No specification without representation!”
Need guidance about how to get involved? E-mail us at firstname.lastname@example.org.