Neuromorphic AI chip provider BrainChip has partnered with engineering team Blue Ridge Envisioneering. Blue Ridge will use the BrainChip Akida processor in its designs for tactical edge devices contracted by defence and intelligence agencies.
The jointly developed devices will be designed for deployment in harsh, resource-restricted environments where power access is limited.
BrainChip’s Akida platform is built to perform effectively in extreme conditions and offers advances edge AI capabilities that makes it optimal for military use cases, the company said.
Stefan Moritz, software engineer at Blue Ridge Envisioneering, said: “Almost all conventional Deep Neural Networks are developed to run on power-hungry GPUs that strain power systems and generate a significant amount of heat; when data must be processed in real-time on a remote device, the challenge for hardware developers is immense.”
The difficulty with utilising deep neural networks stems from their typical need for graphics processing units, which eat up a lot of power. Tactical devices need to be able to process real-time data on a resource budget.
Through the Akida platform, tactical devices will be able to collect data and run advanced AI algorithms without excessive input of power. BrainChip hopes this will make its processor an attractive option for usage in critical situations and decision-making.