Abstract: The growing need for efficient neural network inference in embedded systems has spurred the development of specialized hardware accelerators. This paper introduces the design and ...