At Connect() developer event last year, Microsoft announced the preview of Azure Databricks service for highest-performance streaming analytics projects. Azure Databricks is an Apache Spark-based analytics platform that delivers one-click setup, streamlined workflows and an interactive workspace. Earlier this year, Microsoft announced the general availability of this service.
At the Spark + AI Summit yesterday, Microsoft announced a couple of new enhancements to Azure Databricks.
First, they are adding support for GPU enabled VMs. Developers can now use these VMs to easily build, train and deploy AI models at scale.
Second, they announced a new machine learning runtime that enables distributed, multi-GPU training of deep neural networks using Horovod. Using this new runtime, developers can build deep learning models with a few lines of code.
The runtime also includes HorovodEstimator for seamless integration with Spark DataFrames. Another good news is that it also comes pre-installed and pre-configured with all the necessary packages such as TensorFlow, Keras, and XGBoost. Azure Databricks Runtime for Machine Learning preview is available today as part of premium SKU in Azure Databricks.