A NASA conference paper identified fault-tolerant, highly autonomous computing as necessary for multi-year and permanent lunar habitation. It proposes NASA's High-Performance Spaceflight Computing (HPSC) processor as a radiation-hardened, multicore system-on-chip for lunar and Martian missions, capable of operating without continuous Earth control. HPSC targets more than 100 times the performance per watt of legacy radiation-hardened CPUs and includes autonomous fault detection, diagnosis, and correction.
For an Ark, the central implication is architectural rather than merely computational: habitat systems must continue operating through partial failures, radiation-induced faults, extreme thermal loads, communications loss, and degraded hardware. HPSC's edge-processing capability supports real-time AI, sensor-data ingestion, health management, and graceful degradation, reducing dependence on Earth and limiting single-point failures in life support, power, thermal control, robotics, radiation monitoring, and archival systems. NASA also identifies high-performance onboard computing, extreme-environment avionics, radiation countermeasures, and autonomous health management as civil-space shortfalls.
The Ark team should track HPSC qualification and radiation-testing milestones, currently described by NASA as ongoing in March 2026, and assess HPSC or equivalent architectures for a distributed, heterogeneous compute platform. Research priorities should include cross-generational hardware compatibility, independent compute lanes for safety-critical systems, offline autonomous recovery, radiation-aware memory and storage, formal verification of fault-management software, and long-duration demonstrations that measure availability, graceful degradation, and recovery over decades rather than mission-length timelines.