Intel recently announced the open source release of the deep learning platform Nauta, which provides multiple users to execute the deep learning model in a decentralized computing environment on The Intel processor Xeon. Developers can use the command line interface, web user interface, or the visualization tool TensorBoard. View and monitor the results of model execution, import training data, and developers can use existing datasets, download data online, and create public or private folders to facilitate cross-team collaboration. Recently, companies interested in import of deep learning technology continues to grow, but the deep learning for many companies, is a complicated and time-consuming, and requires psychic integration, validation and optimization model, Intel hopes to borrow from the open source release Nauta deep learning platform, help enterprise developers are more simply deep learning model is set up, for instance, For scalability, the platform uses the components of container tool Kubernetes, with Kubeflow and Docker, to achieve machine learning containerization. At the management level, the platform provides deep learning model templates, eliminating the complex work of establishing and executing experiments. In terms of testing, it also supports batch and stream testing on a single platform. Information source: www.cafes.org.tw/info.asp