Risc V

RISC-V is an open-source instruction set architecture for microprocessors, increasingly used in embedded systems and edge computing due to its flexibility and energy efficiency. Current research focuses on optimizing RISC-V for machine learning workloads, particularly adapting and accelerating neural network inference (including transformer and binary neural networks) through instruction set extensions and hardware-software co-design. This work aims to improve performance and energy efficiency for applications ranging from keyword spotting and image recognition to autonomous navigation in resource-constrained devices, impacting both the development of efficient hardware and the deployment of AI at the edge.