Harnessing Deep Learning and HPC Kernels via High-Level Loop and Tensor Abstractions on CPU Architectures | IEEE Conference Publication | IEEE Xplore