Edge Computing Advances Enable Smarter Spacecraft and Satellites in Orbit

June 1, 2026

As satellites take on more onboard processing — from Earth imaging to autonomy — spacecraft computing designers are pushing for higher performance under tight thermal and radiation constraints. Here’s how suppliers are approaching heat removal, radiation mitigation and production-scale space-grade computing for LEO and beyond.

Design, Manufacturing and Technology Trends

Embedded computing companies developing space-grade components are increasingly highlighting the design challenges of operating in space — environmental constraints the industry has addressed for decades. Thermal management, for example, is critical because the vacuum of space limits heat dissipation primarily to conduction and radiation.

“In space, there’s no air to carry heat away, so thermal management becomes a first-principles problem,” Mark Papermaster, Chief Technology Officer at AMD, writes in his April 2026 blog “AI in Space: Start at the Edge, Build for the Mission.”

“There are a lot of projects going into space. Launches for SpaceX are booked out until 2028, and other launch companies are working as fast as they can to fill the gap. Both commercial and military sectors are moving toward onboard AI and autonomy, shifting away from raw data downlink toward real-time, in-orbit Edge processing and decision making, though some downlink still remains,” said Ralph Grundler, Space Director, Aitech. (Image: Aitech Systems)

“The only way to shed the heat generated by electronics is to conduct it to radiators for heat dissipation. This unique constraint transforms performance-per-watt from a metric into a mandate that drives the architectural innovations making massive-scale AI in orbit a reality,” Papermaster adds.

As on Earth, excess heat can degrade embedded computing performance. Effective thermal approaches will be key as designers equip satellites and spacecraft with more powerful edge-computing devices.

As more companies design embedded hardware for space, radiation effects on electronics remain a core design challenge. As noted in previous Aerospace & Defense Technology (A&DT) coverage of spacecraft computing, radiation in this context refers to ionizing, high-energy cosmic particles.

An overview of embedded computing design challenges, and space applications that are driving demand for more advanced edge computing in space. (Image: Microchip)

Various events can expose electronics to radiation, including solar coronal ejections, particles trapped in Earth’s magnetic fields and galactic cosmic rays — high-energy protons and heavy ions originating from outside our solar system. ¹

Companies take different approaches to mitigating radiation effects on spacecraft electronics. Mercury Systems, for example, differentiates between radiation-hardened and radiation-tolerant electronics. The company notes in a January 2026 blog post that radiation-hardened components are designed for deepspace applications “using specialized materials and processes to resist effects such as total ionizing dose (TID) and single-event effects (SEE).”

In contrast, radiation-tolerant electronics use techniques such as shielding, redundancy or error correction to handle moderate radiation levels. The tradeoff is that radiation-hardened electronics are typically more reliable, but more expensive to develop and sometimes less suitable for advanced, high-performance applications. Radiation-tolerant electronics are often higher performance and less costly, but can be more susceptible to degradation in harsher environments.

After L3Harris Technologies selected Mercury Systems to provide solid-state data recorders (SSDRs) for the U.S. Space Development Agency’s Tranche 3 Tracking Layer constellation, a Mercury representative shared insights with A&DT into how the company developed its radiation-hardened SSDRs for the satellites.

AMD’s Versal system on chip (SoC), pictured here, is being used to enable applications in several new spacecraft and satellites. Blue Origin, for example, is using AMD’s Versal AI Edge Gen 2 adaptive SoCs in its development flight computers that are currently flying in the vehicle testbed that will eventually power the Mark 2 lander that will land astronauts on the moon as early as 2028. NEC also recently announced it is building Japan’s first optical communication satellite constellation with AMD tech. The company will demonstrate high-speed network routing in space using AMD Versal adaptive SoCs to provide high-performance signal processing of data transmissions within the constellation. (Image: AMD)

“Mercury’s unique approach to radiation-hardening is different from traditional shielding-based methods. Our SSDRs are radiation-tolerant by design, utilizing components that are inherently capable of withstanding the harsh radiation environment of space. While shielding may be employed, it serves to enhance existing radiation performance rather than being the primary method of protection,” said Vincent Pribble, Space Product Manager, Mercury Systems. “And Mercury SSDRs do not use any software, eliminating the most likely point of failure and enabling the hardware to run extremely fast. This design-first approach provides more robust and reliable performance throughout the mission life in the challenging space radiation environment.”

Aitech Systems has addressed constant radiation exposure for more than 30 years, spanning more than 5 trillion miles of space flight without error, according to the company’s website.

Ralph Grundler, Space Director, Aitech, described to A&DT how the company uses ruggedization, thermal control and electronics hardening to support the NVIDIA GPU technology it provides for satellites and spacecraft. He also emphasized Aitech’s thermal management techniques.

“Since there is no air in space, systems must rely on cold plate conduction rather than convection cooling. To ensure proper cooling, everything is simulated and tested thermodynamically. Before full testing, Aitech performs detailed thermal simulations to make sure the correct amount and type of materials are used for the best cooling. This is followed by extensive validation, including thermal oven and thermal vacuum (TVAC) testing, to make sure the products will work in space,” Grundler said.

New Technology Drives Increased Manufacturing and Production

Microchip is another supplier with more than 30 years of experience providing embedded processing and other components for space applications. Tim Morin, Senior Director of Microchip Technology’s FPGA Business Unit, provided an overview of the company’s space strategy to A&DT.

Microchip’s RTG4 FPGAs with lead-free, flip-chip bumps have achieved Qualified Manufacturer’s List (QML) Class V and JEDEC qualifications, setting a new standard for mission-critical and modern space applications. (Image: Microchip)

Morin said that the gradual lowering of the cost of launch by Once a space company understands how commercial CPUs and GPUs perform in orbit, it can use that data to design systems with more predictable performance.

Morin also noted that thermal management will remain a design challenge for space-grade computing systems.

“In a vacuum, you can only rely on radiated heat emissions. For most satellites, to overcome that challenge your embedded system typically requires a cold wall to get to a radiator. Some configurations actually attach a radiator to the back of the satellite that points out to deep space to remove heat in the system,” Morin said. “The more compute you have, the more heat you generate, and then the bigger that radiator is. That can increase overall weight and the cost increases because of the increased mass to launch.”

In March, NVIDIA launched its new Space-1 Vera Rubin module. (Image: NVIDIA)

Morin said Microchip addresses thermal constraints by developing radiation-hardened FPGAs that generate up to 50 percent less power than other embedded space computing solutions. In February, Microchip’s RTG4 FPGAs with lead-free, flip-chip bumps achieved Qualified Manufacturer’s List (QML) Class V and JEDEC qualifications, a newer standard aligned to modern space applications.

“We introduced the RTG4 Mil-Plastic family qualified to JEDEC standards and screened through the Radiation-Tolerant (RT) Mil-Plastic flow, which includes full electrical testing across the military temperature range of –55 °C to +125 °C,” Microchip notes in a blog post about the new RTG4 Mil-Plastic. “In the past, the only alternative was Commercial Off-the-Shelf (COTS) devices, but those lacked reliability. RTG4 Mil-Plastic devices provide a cost-effective alternative that uses the same silicon die as QML-qualified parts.”

Vorgao Technologies’ VA41620 microcontroller, pictured here, in a MCU development kit configuration, with a single board computer based on ARM Cortex-M4 processor manufactured with HARDSIL technology and two daughter cards. (Image: Vorago Technologies)

Aitech’s Grundler said LEO applications driving demand for advanced spacecraft computing include Earth imaging, radio and similar missions that collect data requiring onboard processing. While there are discussions about servers or data centers in space, significant challenges remain, so Aitech’s current focus is edge computation and AI processing.

Based on the company’s participation at the recent Space Symposium, Grundler said demand is rising for single-board computers, communications modules, memory, I/O boards and networking hardware such as Ethernet switches — driven by command and data handling, networking and edge processing needs.

“Both commercial and military sectors are moving toward onboard AI and autonomy, shifting away from raw data downlink toward real-time, in-orbit edge processing and decision making, though some downlink still remains,” Grundler said. “Technologically, there is strong momentum toward higher computing using advanced COTS-derived silicon, paired with space-qualified packaging, thermal management and long-term availability. Customers are increasingly looking for integrated, mission-ready computing subsystems rather than individual boards to reduce integration risk.”

The growing number of new spacegrade embedded computing systems introduced over the last year also reflects rising demand for more capable edge computing in orbit. Developers are also placing greater emphasis on lower-cost systems designed specifically for the space environment.

For example, Texas-based semiconductor company VORAGO Technologies recently introduced four radiation-tolerant-by-design (RTbD) microcontrollers featuring ARM Cortex-M4-based cores. The expanded lineup includes the extended-mission VA42620/VA42630 and the cost-optimized VA42628/VA42629 for shorter or lower-orbit missions, reflecting demand for lower-cost, off-the-shelf radiation-hardened computing chips.

Radiation protection is embedded directly into VORAGO’s single-chip microcontroller designs, and the VA4 microcontrollers have also achieved QML-Q+ certification, with VORAGO claiming to be the first “small private chip designer and manufacturer under 100 employees” to achieve QML-Q+.

Other suppliers are expanding manufacturing capability and production capacity in response to rising demand for spacecraft computing systems and components. Ken Hermanny, SVP, Processing Technologies at Mercury Systems, said the company is consolidating some operations and adding automation to reduce delivery times.

“We recently completed the acquisition of specialized manufacturing processes provider SolderMask to further differentiate and accelerate production of our products. We brought together nearly 50 key suppliers to focus on capacity planning, predictability, and performance excellence,” Hermanny said. “We are adding additional shifts in our factories, using AI to better align material supply to demand, and procuring automation tools to improve our output. We also opened a new 50,000 square foot production facility on our Phoenix campus, which will help us ramp production to meet demand for a variety of products, including within our common processing architecture family.”

Citations:

“RAD-HARD Microelectronics for Space Applications,” written by Ross Miller , Director of Marketing, SkyWater Technology, published in the February 2020 edition of Aerospace & Defense Technology.

This article was written by Woodrow Bellamy III, senior editor, SAE Media Group (New York, NY).