TensorFlow* is an end-to-end open source, machine learning framework by Google*. This talk highlights Google and Intel's collaboration on accelerating TensorFlow on x86 CPUs with Intel® oneAPI Deep Neural Network Library (oneDNN) framework. The talk also covers support for new neural-network-focused hardware features such as int8 and bfloat16 vector and matrix extensions that includes a demonstration on select workload speedups.

Speaker

Penporn Koanantakool is a senior software engineer at Google. She leads the TensorFlow performance optimization collaboration with Intel. Penporn holds a PhD in computer science from the University of California, Berkeley, and a BEng in computer engineering from Kasetsart University, Thailand.