More standards, please

Industry trade groups that address standardization issues facing the small form factor and stackables marketplace tend to come in two flavors. Some groups, like the CompactFlash Association, are focused on a single technology and drive that technology across multiple markets through an evolutionary process over time. Other groups, including the SFF-SIG, slice horizontally across a single market (small form factor boards and systems), addressing multiple technologies that might be necessary to drive a complete solution for that market.

Expanding our focus

The SFF-SIG started with a single board-to-board interface standard – SUMIT, or Stackable Unified Module Interface Technology – and quickly realized that, short of endless tweaking and twiddling, we were done with our standardization efforts. We asked ourselves what other technologies constrained the growth of the small form factor market, and whether we had anything to contribute to standardization efforts for these technologies that might solve problems in the SFF space.

As happens with the vast bulk of standardization efforts today, it is rare for a group to start working on a standard with a blank sheet of paper. Every participant brings something to the table and some bring quite a bit – in some cases a fully developed product with documentation ready to be “blessed” by the industry and turned into a standard.

Reviews, risks, and rewards

We’ve been fortunate during the first few years of the SFF-SIG to have some members willing to bring products to the table and subject their efforts to industry peer review as part of the process of standardization. SiliconSystems brought us their SiliconBlade replaceable mass storage technology for standardization (which we call MiniBlade); VIA Embedded brought us the Pico-ITX form factor standard (which we enhanced to Pico-ITXe with the “e” signifying SUMIT expansion along with the definition of Pico-I/O expansion modules); and LiPPERT Embedded Computers brought us the CoreExpress COM standard and contributed significantly to the XR-DIMM Rugged Memory specification with Swissbit and Virtium Technology. Each of these went through three-to-six months of review from SFF-SIG members.

Of course, a strategic shift or company acquisition can orphan a standard, leaving it without a sponsor. In spite of the risks, we’d like to have more standards submitted to SFF-SIG. We don’t believe that continuing standards development in other technologies dilutes our efforts, nor will offering multiple standards to solve the same technology problem compromise our integrity (hence we offer both the COMIT and CoreExpress COM standards). This wider focus can actually have a positive effect. Standards subject to peer review and published by an independent trade group offer a better and more robust long-term solution than standards tied to a single company, even if there are multiple sources for that company’s products.

Moving forward we’d like to offer the SFF-SIG as a home for ’s recently introduced ARM-based COM standard and Advantech’s MIOe interface for SFF mezzanine cards. Publication by an independent industry trade group such as the SFF-SIG goes a long way in building system OEM confidence in the longevity, vendor-neutral control, and breadth of support for these standards. Yes, there are risks when your idea gets in front of a group – somebody’s going to have thoughts about how to make it better. And if you’ve already built one or more products that’s a very uncomfortable discussion, but we can deal with that.

Open forum for open standards

And while we’re at it, we’re open to all kinds of ideas on how to make small systems smaller – thermal solutions, display interfaces, and so on. Bring us your product and let’s create a new standard available to the industry free of charge. Of course, if you are already selling these things, you might have a pretty good head start.

Drop me an email at if you’re interested.