GTC 2020: Speed Up Your Data Science Tasks By a Factor of 100+ Using AzureML and NVIDIA RAPIDS
After clicking “Watch Now” you will be prompted to login or join.
Click “Watch Now” to login or join the NVIDIA Developer Program.
Speed Up Your Data Science Tasks By a Factor of 100+ Using AzureML and NVIDIA RAPIDS
Daniel Schneider, Microsoft | Tom Drabas, Microsoft | Cody Peterson, Microsoft
We'll show how RAPIDS, DASK, and AzureML combine to provide data scientists with access to a parallel high-performance compute environment through a simple Python API. We'll show how the dataframe library cuDF, which will be familiar to Pandas users, as well as the ML library cuML, which provides GPU versions of all ML algorithms available in Scikit-learn, enable data scientists to quickly and easily scale their workloads. Azure Machine Learning service is the first major cloud ML service to integrate NVIDIA RAPIDS, enabling parallel, high-performance computing though a few steps in a simple Jupyter Notebook. RAPIDS is a suite of libraries built on NVIDIA CUDA for doing GPU-accelerated machine learning, enabling faster data preparation and model training. Together, AzureML and RAPIDS dramatically accelerate common data science tasks by leveraging the power of NVIDIA GPUs scaled out to an AzureML cluster of multiple GPU nodes running DASK.