The Deep Learning Virtual Machine (DLVM) is a superset variant of the traditional Azure data science VM which consists of pre-configured environments that are mainly used to develop and deploy deep learning models on top of GPU instances (for example, a Azure GPU NC series VM), and is available on two OSes—Windows Server 2016 and Ubuntu Linux edition.
The DSVM on Azure contains several AI tools that have been pre-built, including CNTK, Keras, Caffe2, and Chainer to pre-process and extract visual data, text, audio, or video data. You can perform data science modelling and use implementation operations by using tools like Microsoft R server, Anaconda Python, Jupyter Notebooks for Python /2.x , R , SQL Server 2017, Azure ML workbench, Julia, F# SDK and so on.
You can provision the Deep Learning VM in the Azure portal...