Uppsala, 11 December 2025 – Rapidity Space and Frontgrade Gaisler are excited to announce their collaboration in the VAIAS project, a strategic initiative aimed at advancing energy-efficient, fault-tolerant AI computing for future space missions. The project is funded through the ESA Phi-Lab Sweden programme, led by RISE Research Institutes of Sweden, with contributions from ESA and Vinnova.
Central to the project is the Frontgrade Gaisler GR801, a next-generation processing platform that uniquely combines a radiation-tolerant NOEL-V RISC-V processor with BrainChip’s Akida neuromorphic processing engine, enabling reliable onboard inference at low power.
Rapidity Space will lead the development of AI and perception capabilities for the GR801 platform. This includes creating a neuromorphic event-based semantic segmentation demonstrator, showcasing how Akida’s architecture can process rapidly changing visual information for applications such as lightning detection, cloud-cover detection, change monitoring, collision avoidance, and rover traversability mapping.
In parallel, Rapidity will extend the LLVM toolchain with improved SIMD/SWAR support and enhanced NOEL-V optimizations, enabling efficient deployment of AI and data-parallel workloads under strict memory and energy constraints. Together, these deliverables broaden the GR801 software ecosystem and demonstrate mission-ready onboard perception for both Earth observation and planetary exploration.