Rhetoric heats up over Small Form Factors (SFF)
BlogApril 10, 2012
Dave Barker
Extreme Engineering Solutions (X-ES)
The debate over Small Form Factor (SFF) design standards is escalating but also becoming entertaining with different views coming from both sides of the Atlantic on how to create SFF designs. Demand for these designs in military C4ISR systems is increasing, but system integrators don?t want to get locked into proprietary solutions. So, embedded computing companies are looking at how they can satisfy this demand as well as the need for lower Size, Weight, and Power (SWaP).
SFF makes sense as the amount of processing power and functionality that is being packed into silicon these days enables a SFF system to do what several years ago would have required several 3U or even 6U boards in a ½ ATR or larger enclosure.
Some companies have looked at defining new SFF standards, such as VITA 73, VITA 74, and VITA 75. Others, like X-ES, are using existing standards, in our case COM Express and PMC/XMC. While other companies have even gone as far as creating a new standards body – the Standardisation Group for embedded Technologies (SGet). SGet was formed by three European companies – Congatec, Kontron and MSC – and they have been joined by Advantech, Data Modul and Seco. It is believed that one of the first SGet workgroups will be looking at the Q Seven form factor.
Since this effort started, there has been some verbal sparring between an unnamed spokesman for SGet, before the organization was officially formed, and Ray Alderman, Executive Director of VITA. I have found this bantering especially entertaining and I hope you do too.
From Ray Alderman’s article in Volume 8, Issue 2 - January 24, 2012 of ConnectorSupplier (www.connectorsupplier.com): They intend to concentrate on board form factors (mechanical dimensions), interfaces (pin-outs on connectors), and integration issues (cables), starting with the Q-7 form factor. This says they are focused on SFF (small form factor) embedded electronics, a commoditized low-margin PC-based electronics board market for benign pedestrian applications. The SFF markets are already terribly fragmented, with over 100 specifications. Are more standards for these small boards truly needed?
From the January, 2012 edition of Micro Technology Europe (http://www.mtemag.com/): “We don’t want companies that don’t participate in one market having control over them,” said the [SGet] spokesman. “Also we don’t want, say, a connector manufacturer dictating what connectors are used on a particular board.” The spokesman said that the aim was to have a different approach to Vita and Picmg. “Vita is trying to look for different fields but it takes them two to three years to come out with a specification, and we see that as too long,” he said. “By the time you have the specifications done, the technology has changed.”
Ray Alderman, ConnectorSupplier: So their claim, that VITA takes too long to complete a standard, is based on their complete lack of engineering knowledge about the complexities of computer architectures. These people don’t have a clue about meshes, stars, double-stars, or switched multiprocessor architectures.
From the January, 2012 edition of MTE: He [the SGet spokesman] also said that there was resentment among European companies that they would do the engineering work on standards and then this would be marketed by American organisations as if the standard came from that country. “We don’t like being dictated to by Vita and Picmg,” he said. “In Europe, we are strong in engineering but not that strong in marketing. We do the work and the US does the marketing and it comes out as a US specification. Why shouldn’t we do it ourselves and be proud of what we do in Europe?”
Ray Alderman, ConnectorSupplier: This new standards group harbors resentment for U.S.-based standards groups. They claim that the U.S. is better at marketing. They further claim that Europeans create wonderful technology, bring it into U.S. standards organizations, and get no credit. The West has been advancing technological innovation for decades. Then it moves east into Europe, as an import. Creating some minor SFF board standards, based on reference designs created in the U.S., does not solve the innovation problem in Europe.
Compared to the advanced high-performance computing architectures engineered at VITA, the standards this new group intends to create are technically unsanitary. But in their increasingly dystopian European markets, it’s probably the best option they could come up with. What they are actually creating is a “landfill technology”: when the boards need replacement or upgrading, you just throw them away.
There is no competition between VITA and this group: They are aiming at the most benign environments with their standards efforts. And, they are aiming at the lowest price/margin segment of the embedded computer industry.
The commodity SFF market looks like an Arkansas trailer park after a tornado, and these guys are just roaming around in the debris.
The embedded computing market hasn’t had any good controversy for a while. While you may not agree that controversy is good for the embedded computing industry, you should agree that it can be entertaining.
