Applications of FPGA in Space

Prepared by Jim Carlquist, VORAGO Technologies, compilation from various sources.
July 11, 2023

Overview
In recent years, the flexibility that Field Programmable Gate Array (FPGA) devices provide has become increasingly important in space applications, where reconfiguration and signal processing capabilities are highly valued. Below, we address some basic questions about FPGAs, as well as specifics about the use of FPGAs in space and how VORAGO Technologies (Austin, Texas USA) and their family of rad hard MCUs can provide industry leading radiation protection to space-based systems. This article is intended as a high-level summary of FAQs on this topic, compiling information from various industry resources and papers. Some sections of this article are directly reproduced from sources cited in the References section at the end of this article.  

The use of reprogrammable FPGAs in space
As the space industry has grown and matured, satellite lifespans have increased exponentially, with some geostationary communications satellites now designed to last multiple decades. Some satellites are now being repurposed or re-sold and reprogrammed for different applications after their initial mission is complete. Coupling these longer lifespans with the enormous cost of building and launching a new satellite, the market is ripe for repurposing of existing in-orbit satellites for a variety of applications. 

The application of FPGAs has moved from simple glue logic to complete subsystem platforms that combine several real time system functions on a single chip, even including microprocessors and memories. The potential for FPGA use in space is steadily increasing, continuously opening up new application areas.

Why is FPGA used in space?

FPGAs are often used in project development because it is simpler to change the design inside the FPGA than to build new hardware.  Development of a custom ASIC can take up to two years, while an SRAM or Flash-based FPGA can be reprogrammed literally dozens of times until the design is perfected.  Space applications demand technology that will limit the risks associated with single-event effects (SEEs) induced by radiation. Certain FPGAs, such as those based on antifuse technology, provide a high degree of immunity against SEEs. Other types of FPGAs may provide some degree of SEE immunity, coupled with better performance, capacity, and reprogrammability, but do not offer the industry leading total ionizing dose (TID) performance of a VORAGO MCU.

Do satellites use FPGA?
Satellites frequently use FPGAs or ASICs (application-specific integrated circuits) as the central processing unit. Over the past 10+ years, FPGAs have become more common and are now replacing ASICs on a regular basis.  The cost of a rad-hard FPGA can be $50,000 or more, so often rad-tolerant or even automotive/industrial grade versions find their way into space.  By adding a true rad-hard MCU, such as the Arm Cortex M4 based VORAGO VA41630, a more complete rad-hard system can be built.  In fact, once an FPGA-based design is complete, the VORAGO VA416xx may be able to replace most or all of that functionality.

FPGA types: SRAM, flash and anti-fuse
There are three major types of FPGA technologies on the market: SRAM-, Flash- and Anti-fuse-based.

SRAM: SRAM-based FPGAs are manufactured in standard CMOS processes, giving the potential for high density devices. SRAM-based FPGAs are the most common type for commercial applications and have the highest capacity and performance of the FPGA technologies discussed here. SRAM-based FPGAs also have the advantage over custom ASIC because they are reprogrammable thus avoiding all development costs associated with ASIC development. However, traditional SRAM-based FPGAs are highly sensitive to ionizing radiation, making them prone to radiation-induced memory upsets. Due to their volatility, SRAM-based FPGAs need to be reprogrammed at power-up.  This limitation creates an unfortunate vulnerability where the FPGA configuration information can be lost.  VORAGO’s MCUs can easily be used as a rad-hard ‘system monitor’ or rad-hard co-processor to prevent this condition.

Flash: Flash-based FPGAs are a non-volatile alternative to SRAM-based FPGAs, based on so-called “floating gates”. The non-volatile nature of flash enables live-on-startup FPGAs without the need for reprogramming, and the flash technology is intrinsically more resistant to radiation compared to SRAM. While flash technology has the advantage of smaller bit storage cells, requiring only one or two transistors to implement a configuration bit storage element compared to the five to six transistors used in SRAM, flash lags behind SRAM in manufacturing process technology. One of the drawbacks of Flash-based FPGA technology is the gradual degradation of configuration memory cells due to charge build-up when reprogramming, limiting the number of times it is possible to reprogram the FPGA. This number, however, is in the order of hundreds of times, and is typically not an issue for space applications. Absorbed radiation over time also leads to charge build-up in the floating gate, eventually rendering the storage cell unusable. This means that flash-based FPGAs in general have a lower acceptable total accumulated radiation dose compared to SRAM-based FPGAs, which is a highly relevant factor for space applications. Also, charge leakage is a problem in flash-based FPGAs, where charge can leak from the floating gate through the insulating material surrounding it resulting in storage cells changing state. This condition, although rare, would be catastrophic.  Adding a VORAGO MCU into the system can reduce the risk and provide years of reliable operation.

Anti-fuse-based: Flash-based FPGAs are more resistant to radiation, as previously mentioned, but include SRAM-based components, mainly in user memory such as D-type Flip-Flops, which are sensitive to upsets. As a third alternative, Anti-fuse-based FPGAs have a distinct advantage in this area. Anti-fuse-based FPGAs have traditionally been used in space applications, and are based on one-time programmable anti-fuse connections. They are less susceptible to radiation-induced errors since the need for configuration bits for each individual interconnect point is eliminated, giving a sort of intrinsic radiation hardening for the configuration. This is also the anti-fuse technology’s greatest disadvantage: once a fuse is “blown” by supplying a large current during programming, it cannot be reprogrammed. This makes antifuse-based FPGAs one-time programmable devices. Anti-fuse FPGAs are also expensive in relation to the performance they offer. These types of devices are less common and do not offer the reprogrammability of the two other types of FPGAs.  

What are the real world applications of FPGA?
FPGA are used in space applications from low earth orbit to deep space, enabling communications, sensors, instruments, and systems for a new generation of space missions and exploration. 

Where are FPGAs most used?
Beyond their use in space, FPGAs are useful for an array of applications in industries such as Aerospace and Defense, Industrial, Communications, Medical, Automotive, and even Consumer Electronics. Coupled with a VORAGO MCU, an FPGA-based system can be built to achieve longer mission performance, higher orbits, and deep space exploration.

Lessons learned from FPGA development
Single-Event Upsets (SEU) induced by radiation have been a limiting factor for the use of FPGAs in space, as SEU may cause involuntary reconfiguration of an FPGA. VORAGO’s family of Arm based microcontrollers can offset these risks.

Suitability of reprogrammable FPGAs in space applications
FPGAs enable high availability computing and connectivity throughput for mission-critical systems in the extremes of space. There are not many radiation-tolerant or rad-hard devices available on the market. Use of a rad-tolerant or rad-hard FPGA gives the system designer the tools to add high speed interfaces and whatever peripherals or connectivity their system requires. Overall chip count is reduced resulting in lower overall size, weight, and power (SWaP). 

Functional Triple Modular Redundancy (FTMR)
Triple Modular Redundancy (TMR) has traditionally been used for protecting digital logic from SEUs in space born applications. The main usage has been either on module level or for the protection of sequential elements in digital logic. With the use of reprogrammable logic, such as Static Random Access Memory (SRAM) based FPGAs, the protection of the sequential logic (with TMR) is insufficient since the logical functionality of the FPGA can be changed due to a charged particle hitting the on-chip configuration SRAM.  Adding one of VORAGO’s family of Arm based microcontrollers can offset these risks.  Protection of the combinatorial logic is therefore required to avoid involuntary changes of functionality.

FPGA-based System-on-Chip (SoC) in space instrumentation
Increasing complexity of satellite and space-based systems requires the addition of an increasing number of cameras, sensors, and other instruments. Connecting to them may require a variety of interface protocols.  Most modern FPGAs contain the available hardware to handle many tasks.  Adding or removing a system component can be more easily done with an FPGA based system.

Radiation-hardened and radiation-tolerant options

Silicon chips are susceptible to the impact of ionizing and non-ionizing radiation. These effects can be classified as either individual, Single Event Effects (SEEs), or Cumulative effects based on Total Ionizing Dose of radiation. Radiation-tolerant FPGAs contain some redundant logic and/or error correcting memories to provide one level of resistance to the harmful effects of both SEE and TID. Rad-hard FPGAs, such as those based on VORAGO’s patented HARDSIL® technology, help to mitigate these effects. HARDSIL is the first line of defense against TID, SEE/SEU, and radiation induced latch-up.  With a HARDSIL-based FPGA, the entire FPGA would be resistant to the effects of both SEE and TID enabling higher orbits and longer duration missions. In addition, VORAGO products are built with dual-interlocking storage cells (DICE) instead of traditional storage cells, implement triple mode redundancy (TMR) on key register elements, and error-correcting (EDAC) memories with periodic ‘scrub’ to correct any single-bit memory errors before they can become multi-bit memory errors. No other rad-hard FPGAs on the market offer this level of radiation protection.

Learn more about VORAGO’s microcontrollers for Aerospace and Defense at: https://www.voragotech.com/aerospace-defense.

Additional Reading:
Dr. Rajan Bedi of Spacechips recently authored this review of spacecraft on-board computing using rad-hard MCUs: https://www.edn.com/spacecraft-on-board-computing-using-rad-hard-arm-mcus/.

References:                                                          
[1] The use of reprogrammable FPGAs in space.

Multiple authors, The European Space Agency.

https://www.esa.int/Enabling_Support/Space_Engineering_Technology/Microelectronics/The_use_of_reprogrammable_FPGAs_in_space#:~:text=%22Field%20Programmable%20Gate%20Array%20(FPGA,(SEU)%20induced%20by%20radiation.

 [2] Suitability of reprogrammable FPGAs in space applications.

Prepared by Sandi Habinc, compilation from various sources. European Space Agency Contract Report by Gaisler Research.

http://microelectronics.esa.int/techno/fpga_002_01-0-4.pdf.

 [3] Functional Triple Modular Redundancy (FTMR): VHDL Design Methodology for Redundancy in Combinatorial and Sequential Logic.

Prepared by Sandi Habinc. European Space Agency Contract Report by Gaisler Research.

http://microelectronics.esa.int/techno/fpga_003_01-0-2.pdf

 [4] SEU Mitigation Techniques for Advanced Reprogrammable FPGA in Space.

Department of Computer Science and Engineering - CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden, FREDRIK BROSSER, EMIL MILH.

http://publications.lib.chalmers.se/records/fulltext/202966/202966.pdf.