From VME to VPX

The New Electronic Packaging Challenge

Sequestration, defense reform and cost constraints are driving a new paradigm in the defense industry, where leveraging existing technologies is no longer optional, but required. This is especially true as we migrate from VME to VPX-based embedded computing.

Figure 1. Technology Readiness Level Chart
The United States government, defense prime contractors, second-tier primes and integrators can no longer afford a “whatever it costs” approach to satisfying technology requirements for deployment. This severely limits the opportunities for developing major new defense electronics platforms. The traditional approach to moving from a Technology Readiness Level (TRL) of 1 to 9 (Figure 1) is too complex and costly in terms of affordability development cycle time, and qualification implications, and is simply no longer acceptable. As such, the defense mission is changing as we approach 2014 and beyond.

Past Practice

Over the past 30 years or more, it has been common for technology migration from TRL level 1 to TRL level 9 to take 20 years or longer. The F-22 project exemplifies this. The initial need for a new fighter to replace the F-16 was outlined in 1981, but the first deployment of the F-22 didn’t occur until 2005. This lengthy transition and development time was due in part to a lack of clear definition of what needed to change. In fact, it wasn’t until 1986 that the initial request for bids went out to prime contractors. In addition, there was no clear definition of how to provide the required capabilities on the platform. This led to the creation of numerous new, interrelated, “custom” technologies. As these new technologies went through the design phase, the impact on other subsystems was in a constant state of flux, leading to changes in the scope of work, which resulted in project delays. These types of changes in leadtime, capabilities, and cost are not acceptable in today’s fast-paced defense industry. Today, these factors are treated as firm fixed targets determined at the beginning of a project.

The focus on affordability, open architecture and rapid deployment of platform upgrades applies pressure to the defense prime contractors to utilize commercially-developed technologies to bring essential upgrades to the theater quickly, and at a reduced cost. This Quick Reaction Capability (QRC) stems from a critical need to apply new technologies in theater as exposures and gaps in systems are identified based on two decades of wartime experience. QRC projects are the baseline of several current product development directions for the military and have significantly shorter cycle time requirements.

Mercury Systems has worked on several mission-critical technology projects designed to affordably move a defense electronics platform from a TRL level 3 to a TRL level 9, in less than 15 months. This rapid development was achieved by utilizing a model that leveraged open standards and proven technology to reduce risk. In this approach, much of the design, development and qualification work is completed in advance and then tailored to each individual project’s needs. By defining the system and subsystem interfaces, commercially- developed products are integrated more easily into subsystems as upgrades at strategic points in the platform’s life cycle. The use of commercially-developed modules that can be used in either lab-based or deployed scenarios, along with commercial subsystems, allows for quick-turn development systems. These lab-based units enable system integrators and primes to begin hardware analysis, as well as software, middleware and firmware development while the final system hardware is still being developed and tested. This “parallel path” concurrent engineering approach drastically reduces development cycle time and the likelihood of design issues cropping up late in the development cycle. The “pinch point” in the system design generally is ad dressed earlier, which significantly reduces costs as well as system development and deployment delays.

Over the years the focus has been on highly defined “fixed” specifications for subsystem solutions that meet SWaP (size, weight and power) constraints. It is apparent today, however, that missions change over time. Today’s platforms must be versatile and easily adaptable to handle a range of ground, sea, and airbased payloads and meet an ever expanding scope of changing mission objectives. This has led to an innovative approach that leverages flexible subsystem modules to introduce new technologies to deployed assets. This modular, or building block, approach allows scalable architectures to be planned into future systems and significantly alleviates the constraints of creating new platforms for future needs.

Figure 2. Module Power Progression
As deployed systems began to migrate from VME to VPX-based embedded computing subsystem solutions, another significant obstacle appeared. While processing speeds and compute density have grown exponentially, the available space, weight and power on the platforms remained constant or was reduced. To that end, the increases in speed and computing density provided opportunities for system consolidation. However, in many cases, the power requirements of a VPX module are two to four times greater than that of a legacy VME module (Figure 2). All of this additional thermal energy must be addressed, without increasing the size or weight of the system.

One solution to these issues is combining flexible open standards, such as VPXbased open architecture, with cutting-edge high reliability packaging solutions. This combination provides the leveragability and capabilities needed in today’s rugged high performance embedded computing market. By embracing scalable open standards- based solutions, the lead time and costs associated with bringing new technologies to market are reduced while scalability and portability are increased. New technologies are changing the potential scope of a project. For example, Mercury’s Air Flow-By™ (AFB) cooling techniques for VITA 48.7/48.1 circuit card assemblies reduce module weight by more than 20 percent, reduce the power of a typical system by greater than 5 percent, and improve the MTBF by five times.

Another trend is the deployment of computational subsystems into a much wider range of uses; for instance, the same subsystem may be deployed on an unmanned aerial vehicle (UAV), a shipbased system, or ground station. The goal is to produce subsystems that are defined by the capability that they provide rather than the platform they were developed for. Accordingly, subsystems must be highly reconfigurable and offer a wide range of modular and scalable capabilities. Although there are a number of challenges in developing these scalable platforms, the product velocity and risk reduction through reuse are attractive incentives for standardizing the computational subsystems.

The demand to deliver cutting-edge technology faster than ever before at lower costs is certainly challenging. Yet, by leveraging open standards and proven technologies in a modular, building block approach, meeting today’s needs with designs that allow for cost-effective, rapid future expansion is a realistic and achievable goal. This approach will ensure that platforms are versatile and adaptable enough to manage diverse ground, sea, and air-based payloads while meeting continuously changing mission goals.

This article was written by Darryl McKenney, Vice President, Engineering Services, and Dan Coolidge, Sr. Mechanical Engineer, Mercury Systems, Inc. (Chelmsford, MA). For more information, Click Here .