Deep Neural Network Library (DNNL)  1.2.2
Performance library for Deep Learning

C++ API example demonstrating how one can perform reduced precision matrix-matrix multiplication using MatMul and the accuracy of the result compared to the floating point computations.

Concepts: