2019-10-15 Intel® Artificial Intelligence Workshop: AI From the Data Center to the Edge (HIAI1W19)

FROM TRAINING TO INFERENCE: CREATING AN END-END DEEP LEARNING PROJECT USING OPTIMIZED SOFTWARE AND HARDWARE FROM INTEL.

Intel® is coming to the Leibniz Supercomputing Centre in Garching near Munich on the 15^th of October with a workshop on Machine Learning and Deep Learning.

In this event, Intel® will be offering an End to End AI workshop, covering topics on Machine Learning, Deep Learning, as well as the Intel AI DevCloud and optimized frameworks and tools. The workshop will be led by the Intel experts, and will walk through cases of how students can use Intel & AI for their own projects.

Topics

Introduction to AI and Intel AI Portfolio

With a quick introduction to AI we delve into the hardware and software offerings from Intel for building Ai solutions.

End to End : From Developing AI model in the Datacenter to Deployment at the Edge (Intel Movidius NCS2)

The first phase of this session gives a handson experience of the different phases involved in training a machine learning model viz. identifying the dataset, cleaning and preparing it,  choosing the right machine learning model, training the model with the right Intel optimization settings etc.

The second phase of this session involves the deployment of the trained model on the Intel Xeon Scalables in the cloud and also an edge deployment usecase enabled by the Intel Movidius NCS.

The workshop is organised by the Leibniz Supercomputing Centre in collaboration with Intel® and the Technical University of Munich.

Please bring your own laptop for the hands-on session!

• An Intel® processor based personal laptop (preferably core series processor from 2015 and beyond) with atleast 8 GB of RAM, 15 GB of free disk space and one or more USB2.0/3.0 port.
• A 64-bit operating system (preferably Windows 10) with a valid installation of Oracle VirtualBox 6 Platform and corresponding Extension Pack.
• Basic knowledge of Python / C++ programming and common Linux commands would be assumed, but not mandatory.