Deep learning, a complex machine learning framework, processes unstructured data like text, video, and images
Intel oneAPI Deep Neural Network Library (oneDNN) improves deep learning framework performance and accelerates application development across numerous hardware architectures
OneDNN is a library of performance functions that offers highly optimized versions of the building blocks for deep learning frameworks and applications
Deep learning building blocks can be found in highly optimized implementations in the Intel oneAPI Deep Neural Network Library (oneDNN)
Utilize hardware- and convolutional parameter-based autonomous memory format selection and propagation
Using Intel Neural Compressor, quantize primitives from FP32 to FP16, bf16, or int8
The oneDNN Graph in PyTorch 2.0 and later speeds up inference on x86-64 CPUs utilising float32 and bfloat16 datatypes