VITA 65 versus OpenVPX: Why can't we all just get along?

We haven't seen this much excitement in the VME community since we went to five-row connectors on P2 back in the late 1980s. No, seriously – I'm not joking. Back then, we still had VME16, we had 3U boards with no useful I/O, and the industry was looking to the future, trying to figure out a way to handle buses wider than 32 bits. The arguments were exceptionally heated, names were called, and predictions called for FutureBus to supplant VME if we so much as dared to convert from three-row DINs.

In the past three weeks alone (from late February to mid-March as I write this), two seemingly friendly camps of VPX supporters have split ranks with one goal in mind: to quickly define interoperable VPX system-level specs leading to interoperable VPX LRUs. The goals are noble and probably above-board for all concerned, but even as this magazine was sent to press, last-minute conference calls were being conducted as "Open Standards Model" supporters squared off against OpenVPX's supporters – with VITA's VSO playing the neutral referee in the ring.

Against OpenVPX

In the interview that precedes this article [see "Mercury's 'OpenVPX Industry Working Group' colors outside VSO's lines: An efficient technology fast-track or Pandora's box?" pg. 12], you can read all about the OpenVPX Industry Working Group. Since that interview was conducted in February, a group of companies consisting of Carlo Gavazzi, Curtiss-Wright Controls Embedded Computing (CWCEC), Elma Bustronic, and X-ES proclaimed its allegiance to open standards administered through the VSO standardization process. In a formal press release entitled "Leading Embedded COTS Vendors Declare Support for Open Standards Model," the group "reiterate[d] and reinforce[d] the importance of maintaining the integrity and openness of the ... VSO," while emphasizing the "spirit of 'coopetition' and openness" in defining an open system-level VPX interoperability specification.

The point of the release? Telling the market that the semi-private OpenVPX approach ain't the way to go; instead, the VPX system spec should be created inside VITA's own VSO – right where VPX started in the first place. It's curious that Elma Bustronic jumped on this bandwagon, since as a box- and backplane-provider, they would win regardless of which approach prevailed. I posed this very question to Justin Moll, director of marketing at Elma Bustronic. He said that his company has provided more VPX backplanes, chassis, and accessories than "all of our competitors combined," and that Elma Bustronic would be "happy to help OpenVPX to bring their ideas into [VSO]." Guess that means they win either way, but prefer doing things in VSO where the rules are well understood.

Enter VITA 65

The issuance of the press release was a mere three days before the first OpenVPX meeting took place prior to the Orlando VSO meeting. Later, during the VSO, four companies proposed the formation of a shiny new working group that would be called VITA 65, "VPX System Specifications and Practices." The goal? To quickly define interoperable VPX system-level specs among LRU vendors. According to CWCEC, the initiative was co-proposed by CWCEC and Elma, along with two other prime contractors. So it now appears that we have two distinct and formal efforts to create an interoperable, system-level specification for VPX: VITA 65 within VSO, and OpenVPX – which is quasi-sanctioned by Ray Alderman of VITA. (Refer to the recent OpenVPX articles listed in our Editor's Note topping off the VPX/VXS Roundtable feature on page 18.)

But of course, Alderman also supports the VITA-centric efforts of Carlo Gavazzi, CWCEC, Elma Bustronic, and X-ES, stating that "VITA welcomes the formation of systems-level standards committees in the VSO. Gathering and comparing documents within, and from outside organizations, the new VSO working group [presumably he's referring to VITA 65 here] can come to consensus on the specification content, within the traditional open-standards process."

Parallel efforts ... cost

In my opinion, VITA is hedging its bets in the interest of speed and defining the best technical specs: Either VITA 65 prevails, or OpenVPX creates a spec and then brings it back into VSO for adoption. Since no one wants two VSO VPX system specs in the long run, both documents will eventually do battle within VITA. I see this costing companies time and money.

This adversarial, parallel effort is what puzzles CWCEC the most; preventing dual efforts is at the heart of the aforementioned press release. According to Mike Hornby, director of marketing at CWCEC, "The VITA spirit is nonexclusionary ... and the goal is to have VITA 65 finished by the Fall [of 2009]." Merely having an OpenVPX document "thrown over the wall" isn't in the best interests of the VPX community or any of the vendors and primes who are closely following these two groups' efforts.

Not to mention the existing design wins. This conundrum is exacerbated by the fact that vendors like CWCEC, X-ES, GE Fanuc Intelligent Platforms, and others are shipping boards into prime contractors. How will those design wins remain stable as all this mayhem unfolds? How can the investments made by all the companies participating in VPX be protected? And equally important: How can primes and vendors invest more time into two separate VPX working groups?

Being "Open, on openness"

CWCEC's Hornby has an answer on what to do with the parallel effort of OpenVPX: "OpenVPX should just shut down." But if the Mercury-sponsored group continues to proceed – as I expect it will – both Elma Bustronic and CWCEC are evaluating ways to work with them.

In reality, both efforts will proceed. OpenVPX will do its thing, as will VITA 65. The irony is that OpenVPX participants can join VITA 65 as long as they're VITA members (and if in "good standing," they can even vote on issues pertaining to VITA 65¹). However, CWCEC was originally excluded from joining OpenVPX. When I queried the Industry Working Group's leadership back in February, they were taking some heat from military primes ("the customers") for excluding CWCEC and seemed ready to open the door to CWCEC, subject to terms such as a nondisclosure agreement and memorandum of understanding. At the time of this writing, CWCEC had not joined OpenVPX, and Elma Bustronic wouldn't comment on the record.

But CWCEC's Hornby and Elma Bustronic's Moll argue that the reintegration of OpenVPX's system spec back into VSO can be hastened by VITA 65's "openly and proactively report[ing] progress and significant achievement milestones to VITA members at large." Presumably this will make both groups' efforts synergistic and easier to merge later in 2009.

We'll see. CS

¹ OpenSystems Media, publisher of VME and Critical Systems magazine, is a VITA member. Facts reported in this article represent either public information or data gleaned directly from interviews.