Is NASA's design opportunity for FPGAs in space vanishing in favor of privatized platforms?

-July 10, 2013

Not so long ago, the opportunities for rad-hardened FPGAs used in space applications rested with the Defense Department or NASA. The rise of privatized launches and open-architecture microsatellites like CubeSat, however, have made NASA design-ins the exception rather than the rule.

This reality was drilled home in late June as 4DSP LLC announced a $42,479 contract from NASA's Langley Research Center, to use 3U CompactPCI cards based on a Virtex-6 as part of a terrestrial platform to test space instruments. Only a few years ago, NASA and Air Force contracts utilizing FPGAs were commonplace. Now, it's time for microsatellites designed by academia and private industry.

There is not a strict delineation between the two realms. NASA, the National Reconnaissance Office, and the Air Force are experimenting with designs like CubeSat and FalconSat. But government contracts have long relied on proprietary satellite buses such as A2100 and STARbus. Open, simplified platforms have had to challenge high-value architectures promoted by gigantic aerospace corporations. Thus, startups and coalitions in the privatized launch and satellite realms have been able to move to new bus and module architectures with less concern for legacy architectures.

We cited one example of the popular new CubeSat at the end of last year, when the University of Michigan announced an MCube2 satellite design based on Xilinx FPGAs. Since that time, the floodgates for FPGAs in the CubeSat community have opened.

Here's just a few examples of what we are seeing:
  • COSMIAC (Configurable Space Microsystems Innovations & Applications Center) in Albuquerque is offering designers a Spartan-based set of boards for software-defined radio and other space applications.
  • A University of Florida team is working on energy-budget allocations for the Virtex-4QF used in CubeSat environments.
  • The EADS Astrium payload for the UKube 1 CubeSat in the UK will be based on a Virtex-4 architecture.
  • NASA itself is getting involved in a Virtex-5QV on-board processor for CubeSat called COVE (CubeSat On-board processing Validation Experiment). NASA's Jet Propulsion Laboratory is working with several private and academic groups on improving core processing for CubeSat platforms.

The last example shows that NASA need not work at cross-purposes to emerging independent efforts to build microsatellites. Rather, collaborations with schools and startups may allow NASA to conduct core research more cheaply, while exploring cost-effective alternatives to government launch platforms, such as rockets from Space X.  In the new world, FPGAs may have more opportunity to end up in orbit, because designs need not be limited to strapped federal agencies currently facing sequestration.

Also see:

Loading comments...

Write a Comment

To comment please Log In