Start Date: 16/05/2022

/ 14 Tcs

DURATION 2 Days

Course Overview

Course Description

This course describes how to use the Vitis™ AI development platform in conjunction with DNN algorithms, models, inference and training, and frameworks on cloud and edge computing platforms.

The emphasis of this course is on:

▪Illustrating the Vitis AI tool flow

▪Utilizing the architectural features of the Deep Learning ProcessorUnit (DPU)

▪Optimizing a model using the AI quantizer and AI compiler

▪Utilizing the Vitis AI Library to optimize pre-processing andpost-processing functions

▪Creating a custom platform and application

▪Deploying a designLevel – AI 3 Course Details days ILT 2Course Part Number – AI-INFEWho Should Attend? – Software and hardware developers, AI/ML engineers, data scientists, and anyone who needs to accelerate their software applications using Xilinx devices Prerequisites ▪Basic knowledge of machine learning concepts▪Neural Networks Explained – Machine Learning Tutorial forBeginners -www.youtube.com/watch?reload=9&v=GvQwE2OhL8I▪How Convolutional Neural Networks Work -www.youtube.com/watch?v=FmpDIaiMIeA▪Comfort with the C/C++/Python programming language▪Software development flowSoftware Tools▪Vitis AI development environment 1.1▪Vivado Design Suite 2019.2Hardware▪Architecture: Xilinx Alveo™ accelerator cards, Xilinx SoCs, andACAPs

After completing this comprehensive training, you will have the necessary skills to:

▪Describe Xilinx machine learning solutions with the Vitis AIdevelopment environment

▪Describe the supported frameworks, network modes, andpre-trained models for cloud and edge applications

▪Utilize DNN algorithms, models, inference and training, andframeworks on cloud and edge computing platforms

▪Use the Vitis AI quantizer and AI compiler to optimize a trainedmodel

▪Use the architectural features of the DPU processing engine tooptimize a model for an edge application

▪Identify the high-level libraries and APIs that come with the XilinxVitis AI Library

▪Create a custom hardware overlay based on applicationrequirements

▪Create a custom application using a custom hardware overlayand deploy the design

Course Outline:

Course Outline
Day 1
▪Introduction to the Vitis AI Development Environment
Describes the Vitis AI development environment, which consistsof the Vitis AI development kit, for AI inference on Xilinx hardwareplatforms, including both edge devices and Alveo acceleratorcards. {Lecture}
▪Overview of ML Concepts
Overview of ML concepts such as DNN algorithms, models,inference and training, and frameworks. {Lecture}
▪Frameworks Supported by the Vitis AI DevelopmentEnvironment
Discusses the support for many common machine learningframeworks such as Caffe and TensorFlow. {Lecture}
▪Setting Up the Vitis AI Development Environment
Demonstrates the steps to set up a host machine for developingand running AI inference applications on cloud or embeddeddevices. {Demo}
▪AI Optimizer
Describes the optimization of a trained model that can prune amodel up to 90%.
This topic is for advanced users and will be covered in detail inthe Advanced ML training course. {Lecture}
▪AI Quantizer and AI Compiler
Describes the AI quantizer, which supports model quantization,calibration, and fine tuning. Also describes the AI compiler toolflow.
With these tools, deep learning algorithms can deploy in the DeepLearning Processor Unit (DPU), which is an efficient hardwareplatform running on a Xilinx FPGA or SoC. {Lecture, Lab}
▪AI Profiler and AI Debugger
Describes the AI profiler, which provides layer-by-layer analysis tohelp with bottlenecks. Also covers debugging the DPU runningresult. {Lecture}
▪Introduction to the Deep Learning Processor Unit (DPU)
Describes the Deep Learning Processor Unit (DPU) and itsvariants for edge and cloud applications. {Lecture}
▪DPU-V1 Architecture Overview
Overview of the DPU-V1 architecture, supported CNN operations,and design considerations. {Lecture}
▪DPU-V2 Architecture Overview
Overview of the DPU-V2 architecture, supported CNN operations,DPU data flow, and design considerations. {Lecture}
Day 2
▪Vitis AI Library
Reviews the Vitis AI Library, which is a set of high-level librariesand APIs built for efficient AI inference with the DPU. It providesan easy-to-use and unified interface for encapsulating manyefficient and high-quality neural networks. {Lecture, Lab} [Notethat this lab is not available in the OnDemand version as anevaluation board is required for the entirety of the lab]

▪Creating a Custom Hardware Platform Using the VivadoDesign Suite Flow (Edge)

Describes the steps to build a Vivado Design Suite project, addthe DPU-V2 IP, and run the design on a target board. {Lab}

▪Creating a Custom Application (Coming Soon)

Illustrates the steps to create a custom application, such asbuilding the Linux image, optimizing the trained model, and usingthe optimized model to accelerate the design. {Lecture, Lab}

▪Creating a Custom Hardware Platform Using the VitisEnvironment Flow (Edge)

Describes the steps to build a Vitis unified software platformproject that adds the DPU as the kernel (hardware accelerator)and to run the design on a target board. {Lab}