VPX and PC/104: Coexisting to Support Different Military Applications

1 of Enlarge image

Parvus provides a range of rugged VPX and PC/104 platforms

In the world of military electronics, SWAP (size, weight and power) rules supreme. Unmanned and manned applications alike are often demanding the highest computational performance possible squeezed into highly dense, low-power solutions. There is growing demand for embedded subsystems capable of handling distributed multiprocessing architectures, serial switched fabrics, and enormous amounts of sensor I/O. These increasingly demanding technical requirements have sparked the development of numerous small form factor board architectures, including the meteorically rising VPX.

Aimed squarely at network-centric defense applications, VPX brings the performance of switched fabric interconnects such as PCI Express, Serial RapidIO and 10 Gigabit Ethernet into a rugged slot-card architecture familiar to legacy VME users. VPX affords SWAP advantages as users transition from VME to VPX’s high density, high bandwidth I/O and 3U form factor. Its conduction cooled formats also support blade-like, field replaceability (2 Level maintenance) with enhanced ESD/dust protection.

While the parallel VME bus architecture ranks in predominance in the military/aerospace market – thanks to its legacy installed base – VPX is poised to become the successor of traditional VME for a growing number of rugged applications. VPX may still be in its infancy; yet accordingly to industry estimates, it already claims more than 20% of the market previously served by VME. The rapid adoption of VPX is further accelerating due to the interoperability defined in the OpenVPX standard and the growing ecosystem of suppliers adopting it.

Different Boards for Different Applications

So while its ruggedness and performance capabilities point to VPX as the savior of military electronics, it is certainly not a catchall solution for every military application. In fact, for some, VPX systems may be overkill — too complex, too large, too expensive, or just too hot to handle. There are certainly tradeoffs. One size does not fit all.

For example, VPX uses more power and runs hotter than legacy buses like VME, PC/104 and CompactPCI. It is not uncommon for VPX cards to run at 100+ Watts per slot. VPX has plenty of power pins (enough to theoretically support up to 768W/slot) yet it presents a significant thermal management challenge to deal with this level of computational density. Sophisticated and highly effective cooling techniques, such as spray or liquid cooling are being implemented, yet this also adds to the complexity for system design and maintenance compared to passive cooling methods. Consequently, cost sensitive applications, those with lower computational needs and battery-powered systems may not be well suited for the VPX architecture.