VLSI Structure-aware Placement for Convolutional Neural Network Accelerator Units | IEEE Conference Publication | IEEE Xplore