The most transformative science increasingly lives at the intersection of physics-based simulation and AI — where machine learning surrogates accelerate multiscale models, generative AI proposes candidate molecules or materials for high-fidelity validation, and traditional solvers generate the training data that makes foundation models scientifically trustworthy. Meeting the computational demands of these converged workflows requires infrastructure that excels at both: dense numerical simulation and large-scale AI, on the same platform, within the same workflow. The Vera Rubin architecture and NVIDIA's full accelerated computing stack are designed for exactly this convergence, powering the world's leading supercomputing centres and AI Factories. This talk explores how NVIDIA's platform — from silicon to software libraries to application frameworks — enables researchers to fluidly combine simulation, AI training, and inference at scale, compressing discovery timelines from years to days across climate science, drug design, materials engineering, and beyond.
