Andreea Munteanu, MLOps Product Manager, Canonical; Michael Balint, Sr Manager, Product Architecture, NVIDIA
High-performance AI can only be achieved when hardware and software work together. An end-to-end MLOps solution working on hardware designed for AI is crucial for enterprises wanting to move their AI out of their labs and scale it to production. Did you know that Charmed Kubeflow on NVIDIA hardware enables you to take AI models from concept to production rapidly, robustly and reliably?
Although the platform can be used in any sector, this webinar will focus on life sciences. While a stack like this is suitable for various use cases across the life sciences vertical, we specifically focus on image analysis, a key activity in the drug discovery pipeline. Different stages, such as monitoring disease progression or structure determination of biological macromolecules, benefit from advances in deep learning which let you analyse large volumes of data.
During the webinar, NVIDIA and Canonical’s experts will cover:
NVIDIA DGX, networking, and the requisite hardware for AI training
Image analysis using deep learning for drug discovery
Challenges to performing image analysis
NVIDIA SDKs and AI frameworks via NGC
From experimentation to production: the challenge to maintain and monitor various models
How to optimise operations for AI at scale
Solutions for deep learning: hardware and software combined together