BEGIN:VCALENDAR VERSION:2.0 PRODID:icalendar-ruby CALSCALE:GREGORIAN BEGIN:VEVENT DTSTAMP:20260706T192715Z UID:f528670a-fbfa-4aad-a3c5-43514512281d DTSTART:20200615T080000Z DTEND:20200618T140000Z DESCRIPTION:This course will be delivered as an ONLINE COURSE for remote p articipation because of the COVID-19 measures enforced by most European go vernments.\n\nREGISTRATION is strictly NECESSARY since the details to ac cess the online course will be provided to the registered and accepted att endees only.\n\nThe workshop will take place online 10:00-12:00 and 13:00- 16:00 CEST each day.\n\nIf you want to be set on the waiting list\, please contact weinberg@lrz.de.\n\nOverview\n\nLearn how to train and deploy a n eural network to solve real-world problems\, how to generate effective des criptions of content within images and video clips\, how to effectively pa rallelize training of deep neural networks on Multi-GPUs and how to accele rate your applications with CUDA C/C++ and OpenACC.\n\nThis 4-days worksho p  combines lectures about fundamentals of Deep Learning for Multiple Dat a Types and Multi-GPUs with lectures about Accelerated Computing with CUDA C/C++ and OpenACC.\n\nThe lectures are interleaved with many hands-on ses sions using Jupyter Notebooks. The exercises will be done on a fully confi gured GPU-accelerated workstation in the cloud.\n\nThe workshop is co-orga nized by LRZ\, IT4Innovations and NVIDIA Deep Learning Institute (DLI) for the Partnership for Advanced Computing in Europe (PRACE). LRZ as part of GCS and IT4Innovations are both PRACE Training Centres which serve as Euro pean hubs and key drivers of advanced high-quality training for researcher s working in the computational sciences.\n\nNVIDIA DLI offers hands-on tra ining for developers\, data scientists\, and researchers looking to solve challenging problems with deep learning.\n\n All instructors are NVIDIA c ertified University Ambassadors.\n\nAgenda\n\n1st day: Fundamentals of Dee p Learning for Multiple Data Types\n\nThis day explores how convolutional and recurrent neural networks can be combined to generate effective descri ptions of content within images and video clips.\n\nLearn how to train a n etwork using TensorFlow and the Microsoft Common Objects in Context (COCO) dataset to generate captions from images and video by:\n\n\n Implementing deep learning workflows like image segmentation and text generation\n Com paring and contrasting data types\, workflows\, and frameworks\n Combining computer vision and natural language processing\n\n\nUpon completion\, yo u’ll be able to solve deep learning problems that require multiple types of data inputs.\n\n2nd day: Fundamentals of Accelerated Computing with Op enACC\n\nOn the last day you learn the basics of OpenACC\, a high-level pr ogramming language for programming on GPUs. Discover how to accelerate the performance of your applications beyond the limits of CPU-only programmin g with simple pragmas. You’ll learn:\n\n\n \n How to profile and optimiz e your CPU-only applications to identify hot spots for acceleration\n \n \ n How to use OpenACC directives to GPU accelerate your codebase\n \n \n Ho w to optimize data movement between the CPU and GPU accelerator\n \n\n\nUp on completion\, you'll be ready to use OpenACC to GPU accelerate CPU-only applications.\n\n3rd day: Fundamentals of Accelerated Computing with CUDA C/C++\n\nThe CUDA computing platform enables the acceleration of CPU-only applications to run on the world’s fastest massively parallel GPUs. On t he 3rd day you experience C/C++ application acceleration by:\n\n\n \n Acce lerating CPU-only applications to run their latent parallelism on GPUs\n \ n \n Utilizing essential CUDA memory management techniques to optimize acc elerated applications\n \n \n Exposing accelerated application potential f or concurrency and exploiting it with CUDA streams\n \n \n Leveraging comm and line and visual profiling to guide and check your work\n \n\n\nUpon co mpletion\, you’ll be able to accelerate and optimize existing C/C++ CPU- only applications using the most essential CUDA tools and techniques. You ’ll understand an iterative style of CUDA development that will allow yo u to ship accelerated applications fast.\n\n4th day: Fundamentals of Deep Learning for Multi-GPUs\n\nThe computational requirements of deep neural n etworks used to enable AI applications like self-driving cars are enormous . A single training cycle can take weeks on a single GPU or even years for larger datasets like those used in self-driving car research. Using multi ple GPUs for deep learning can significantly shorten the time required to train lots of data\, making solving complex problems with deep learning fe asible.\n\nOn the 2nd day we will teach you how to use multiple GPUs to tr ain neural networks. You'll learn:\n\n\n \n Approaches to multi-GPUs train ing\n \n \n Algorithmic and engineering challenges to large-scale training \n \n \n Key techniques used to overcome the challenges mentioned above\n \n\n\nUpon completion\, you'll be able to effectively parallelize training of deep neural networks using TensorFlow.\n\nImportant information\n\nAft er you are accepted\, please create an account under courses.nvidia.com/jo in .\n\nEnsure your laptop / PC will run smoothly by going to http://webso cketstest.com/ Make sure that WebSockets work for you by seeing under Envi ronment\, WebSockets is supported and Data Receive\, Send and Echo Test al l check Yes under WebSockets (Port 80).If there are issues with WebSockets \, try updating your browser. If you have any questions\, please contact M arjut Dieringer at mdieringer"at"nvidia.com. \n\nPRACE Training and Educa tion\n\nThe mission of PRACE (Partnership for Advanced Computing in Europe ) is to enable high-impact scientific discovery and engineering research a nd development across all disciplines to enhance European competitiveness for the benefit of society.  PRACE has an extensive education and trainin g effort through seasonal schools\, workshops and scientific and industria l seminars throughout Europe. Seasonal Schools target broad HPC audiences\ , whereas workshops are focused on particular technologies\, tools or disc iplines or research areas.\n\nNVIDIA Deep Learning Institute\n\nThe NVIDIA Deep Learning Institute delivers hands-on training for developers\, data scientists\, and engineers. The program is designed to help you get starte d with training\, optimizing\, and deploying neural networks to solve real -world problems across diverse industries such as self-driving cars\, heal thcare\, online services\, and robotics.\n\n\nhttps://events.prace-ri.eu/e vent/1007/ SUMMARY:[ONLINE] Deep Learning and GPU programming workshop @ LRZ URL;VALUE=URI:https://events.prace-ri.eu/event/1007/ END:VEVENT END:VCALENDAR