Automatic Generation of High-Performance Inference Kernels for Graph Neural Networks on Multi-Core Systems Conference Paper uri icon