MEGA: A Memory-Efficient GNN Accelerator Exploiting Degree-Aware Mixed-Precision Quantization | IEEE Conference Publication | IEEE Xplore