Data Pipeline Development: Increasing Data Science Productivity and Performance

The data science workflow has traditionally been slow and cumbersome when it comes to loading, filtering, and manipulating data, as well as ML training itself. These processes were constrained to slow, CPU-based computing, and resulted in lengthy cycle times impacting data science productivity. NVIDIA RAPIDS delivers GPU-accelerated machine learning and analytics libraries, deployed on NVIDIA GPU-platforms for maximized data science productivity, performance, and insights.

Join us for this live webinar to learn:

• How RAPIDS accelerates your Python data science tool-chain with minimal code changes and no new tools to learn
• The use of the CUDA Array Interface to exchange data between GPU-accelerated libraries, including deep learning frameworks
• Jupiter notebook examples that demonstrate the usage of these libraries