BrainChip, Gaisler work on AI in space
BrainChip has said it is working to integrate its Akida neuromorphic processor IP into Frontgrade Gaisler’s fault-tolerant, radiation-hardened microprocessors.
The collaboration is set to provide a first space-grade system-on-chip with AI capability, BrainChip said. The inclusion of Akida would provide improved power efficiency and inference performance for soft-AI while maintaining compatibility with existing Convolutional Neural Networks (CNNs), the company said.
The companies did not indicate how quickly the integration would take place or when Frontgrade Gaisler would be able to offer the capability. Nor did the company say whether Akida would be added to both the SPARC and RISC-V product lines or just one. Gaisler offers processors and systems-on-chip (SoC) based on the SPARC architecture (Leon) and on the RISC-V architecture (Noel).
“Adding AI capabilities to our next generation microprocessors would set a new standard for modern space-grade computing devices,” said Sandi Habinc, general manager at Frontgrade Gaisler, in a statement issued by BrainChip. “The success of this joint effort could empower organizations to leverage AI technology to enhance mission efficiency, ultimately advancing the frontiers of what is possible in space.”
Ali Zadeh, head of microelectronics at the European Space Agency, said: “Integrating neuromorphic capabilities in a space-grade SoC represents an exciting technological avenue for the next generation of space applications.”
Related links and articles:
News articles:
BrainChip strenthgens IP portfolio with event-based pattern detection patent
Imsys develops RISC-V core, looks to AI in space
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
