The Case for In-Memory Inferencing of Quantized CNNs at the Edge

As artificial intelligence becomes a pervasive tool for the billions of IoT devices at the edge, the data movement bottleneck imposes severe limitations on these systems’ performance and autonomy. Processing-in-Memory emerges as a way to mitigate the data movement bottleneck while satisfying the stringent performance, energy efficiency, and accuracy requirements of edge imaging applications that rely on convolutional neural networks.