OIT’s Research and Data Science Services Team, in coordination with NVIDIA and Center for Research Computing, is offering two new free NVIDIA DLI workshops to the SMU community. These all-day workshops will provide the foundations for using multiple GPUs on a single computer with Nvidia CUDA and how to speed up data engineering tasks using GPUs. In addition, those who successfully complete the workshop assessment will receive an NVIDIA DLI certificate.
All workshops are in Ford Hall 124 from 9 a.m. to 5 p.m. A personal laptop and registration is required. Lunch will be provided.
NIVIDA DLI: Fundamentals of Deep Learning for Multi-GPUs
Learn some of the basic building blocks of machine learning. In this workshop we will learn how to use Horovod to scale a model running on a single GPU to multiple GPUs. We will be using prebuilt models and explaining some of the components, so prior machine learning experience is helpful but not required to complete this workshop.
NVIDIA DLI: Accelerating Data Engineering Pipelines
Learn how to speed up data engineering tasks using GPUs. In this workshop we will discuss ways to manipulate large sets of data quickly using tools like Pandas, cuDF, Dask, and NVTabular. We will use those techniques to make a Plotly dashboard to display large data sets quickly.
Don’t forget, the Research and Data Science Services Team, in coordination with the Center for Research Computing and the Data Science Institute, is offering more free workshops to the SMU community. Find out more in Research and Data Science Services Offer More Free Workshops this Fall.
For more information or to request a consultation, feel free to email firstname.lastname@example.org.
Update 10/05/2022: Program change. Fundamentals of Deep Learning for Multi-GPUs replaces previously scheduled workshop on October 10, 2022.