Distributed Deep Learning on Apache Spark w/ BigDL - incl Caffe/Torch/TensorF­low

Intel recently released BigDL, an open source distributed deep Learning framework for Apache Spark ( https://github.com/intel-analytics/BigDLhttp://software.intel.com/bigd ). It brings native support for deep learning functionalities to Spark, provides orders of magnitude speedup than out-of-box open source DL frameworks (e.g., Caffe/Torch/TensorFlow) with respect to single node Xeon performance, and efficiently scales out deep learning workloads based on the Spark architecture. In addition, it also allows data scientists to perform distributed deep learning analysis on big data using the familiar tools including python, notebook, etc.  In this talk, we will give an introduction to BigDL, show how big data users and data scientist can leverage BigDL for their deep learning (such as image recognition, object detection, NLP, etc.).  BigDL users can analyze large amounts of data in a distributed fashion, which allows them to use their big data (e.g., Apache Hadoop and Spark) cluster as the unified data analytics platform for data storage, data processing and mining, feature engineering, traditional (non-deep) machine learning, and deep learning workloads.

Speaker: Sergey Ermolin, Intel