FPGAs have already changed the cost/reliability paradigm for embedded systems in high-reliability applications, thanks to advances in hardness and power reduction. But on many embedded applications for high-reliability markets, designers depend on a number of peripheral analog components such as analog-to-digital and digital-to-analog converters to talk to the real world. Other system components such as phase-locked loops (PLLs) and DC/DC converters are usually required to complete a system design. These peripherals impact overall cost, size and reliability. Peripheral analog parts can also be challenging to work with and to source for radiation environments, as an example.
To further leverage the power of FPGAs, military-and-aerospace engineers are actively looking for ways to integrate many of these analog functions onto the FPGA. Synthesizable, digital IP cores that replace some analog functions now exist, allowing mil/aero designers to implement ADC, DAC, DC/DC controller and clock-multiplier functions in fully digital processes such as FPGAs. Not only does this new ability leverage the advantages of FPGAs, it also helps mitigate many challenges of using analog components in high-reliability applications.
Overcoming high-reliability design challenges
The engineering challenges of designing for military or high-reliability applications such as aerospace are numerous. Power and weight are usually under strict budgets because they can affect operating costs and insertion costs exponentially. Physical shock safeguards, force survival and protection from single-event upsets (SEUs) and latchup often mean that parts are larger, heavier and more power hungry than commercial devices. For instance, a commercial 12-bit, 10-MHz bandwidth ADC measures approximately .71 by .42 inches and consumes 280 milliwatts. The equivalent radiation-hardened part is .81 by .72 inches and consumes 335 mW. That’s almost double the size at 20 percent more power.
A wide temperature range is another issue. Typically, temperatures of -40°C to +80°C are expected for many military embedded applications here on Earth. Temperature takes on another complexion in space. In satellite electronics design, for instance, the normal operating junction temperature might be -55°C to +125°C. Monitoring this onboard temperature is key to effective system maintenance, but installing a rad-hard ADC part to provide this function can add up to one square inch of board and require additional components and testing.
When a high-reliability design makes use of peripherals such as ADCs, DACs, DC/DC converters or PLLs, each one of those components represents a possible point of failure. Each must be qualified and tested, and each is most likely not optimally designed for the specific need. There is also always a risk that the manufacturer will discontinue the part, forcing requalification of the entire system.
These challenges to working with analog components in high-reliability environments can evaporate by using the FPGA for a unified, all-digital approach. Let’s take a look at this new paradigm in military/aerospace design.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.