Enabling access to NVIDIA compute acceleration is an key component of VMware's approach to enabling HPC and ML workloads on vSphere. In this talk we will discuss the various available options, provide performance results, and share some hardware and software tips and guidance to help you meet the needs of your organization's data scientists and researchers.
In the first part of this workshop, you will learn about the concepts of machine learning (ML), including inference and training, supervised and unsupervised learning, and deep learning. Bring your laptop to build a neural network model yourself. Gain an understanding of how ML works and what types of problems can be solved with these techniques. You will also see a demo of a data science automation tool to illustrate how ML is used by data scientists. The second part of the workshop will focus on the IT infrastructure needed to support ML, including a detailed discussion of the ways specialized hardware accelerators such as GPUs are enabled on VMware vSphere, along with a discussion of ML performance. Learn about the role of Apache Spark and Hadoop in supporting machine learning.