"Frontgrade Gaisler secures astronautical AI technology from BrainChip through Akida IP licensing"
The European Space Agency (ESA) is leading the charge in deploying neuromorphic computing for space missions, aiming to improve edge deployments in space and push the boundaries of artificial intelligence (AI) to the furthest reaches.
According to Laurent Hili, a microelectronics and data handling engineer at the ESA, neuromorphic computing is essential for computer vision applications in space. The ESA's program has demonstrated the effectiveness of neuromorphic technology through collaborations with satellite primes and IP/components suppliers.
One such supplier is BrainChip Holdings, whose neuromorphic AI has been chosen for its real-time data stream processing, resiliency, autonomy, and accuracy at low power, mass, and volume. BrainChip's Akida neuromorphic computing solution is designed to enhance space-borne, on-board computers, offering superior power efficiency and inference performance.
The Akida solution maintains compatibility with existing Convolutional Neural Networks (CNNs), making it an attractive choice for the ESA. BrainChip Holdings has licensed its Akida IP to Frontgrade Gaisler for space-grade, fault-tolerant system-on-chip solutions.
Frontgrade Gaisler, another key player in the ESA's program, will leverage BrainChip's technology to develop advanced space-grade electronics. The goal of this collaboration is to deploy AI chips in space, with the first applications expected to launch in 2024.
The license agreement between the ESA, BrainChip Holdings, and Frontgrade Gaisler marks a significant step forward in the deployment of neuromorphic AI in space. Current FPGA or GPU technology does not bring satisfactory results for edge deployments considering mass, volume, and power constraints, according to Laurent Hili.
As the ESA continues to deploy neuromorphic computing for space missions, it is clear that this technology will play a crucial role in the future of space exploration and AI. The ESA's program has demonstrated the superiority of neuromorphic technology through several ongoing activities with satellite primes and IP/components suppliers such as Frontgrade Gaisler and BrainChip.
In the coming years, we can expect to see more advancements in neuromorphic AI for space applications, as the ESA leads the efforts to deploy neuromorphic computing to the furthest edge of AI.
