We often think of memories like the contents of a museum: static exhibits that we view to understand the present and prepare ...
Abstract: Artificial intelligence (AI) edge devices increasingly require the enhanced accuracy of floating-point (FP) multiply-and-accumulate (MAC) operations as well as nonvolatile on-chip memory to ...
Abstract: We propose an embedded Neural Processing Unit (NPU) architecture for deep learning inference to address the stringent energy, area, and cost requirements of edge AI. This heterogeneous ...