MapR Adds TensorFlow to Customer 360 Quick Start Solution
March 2017 by Emmanuelle Lamandé
MapR Technologies, Inc., announced the availability of its enhanced Customer 360 Quick Start Solution featuring TensorFlow powered by the MapR Converged Data Platform.
TensorFlow is an open source software library for machine learning capabilities, and is used for diverse use cases including language translation and medical diagnosis. Combining TensorFlow with MapR enables customers to build converged applications that incorporate deep learning and natural language processing to better understand customer behavior and uncover deeper insights in a 360-degree-customer-view solution.
TensorFlow applications powered by MapR can leverage petabytes of data and train models in parallel across multiple servers, leveraging both CPUs and/or GPUs. TensorFlow runs on MapR clusters with heterogeneous hardware for fast access to multi-temperature web-scale storage and support for other compute engines such as Apache Spark in order to build robust data science pipelines.
TensorFlow benefits from platform-level capabilities in MapR, including microservices support to enable easier management of input and output data, particularly for A-B/multivariate testing of machine learning models. In addition, MapR support for data persistence in containers provides another framework for deploying TensorFlow jobs.
The new MapR Customer 360 Quick Start Solution with TensorFlow is available today from MapR worldwide.