10-749 AI for Scientific Computing
12 units, Carnegie Mellon University
AI for Scientific Computing introduces emerging techniques at the intersection of AI, machine learning, and applied mathematics for solving computational problems in the sciences. The course is technical in nature and covers the mathematical foundations of these methods, practical engineering details, benchmarking of current approaches, and obstacles to deployment in real-world scientific settings. Topics include physics-informed neural networks (PINNs), PDE solvers and surrogate models such as neural operators, with applications to computational fluid dynamics and climate; discrete and continuous diffusion models, with applications to protein design and nonequilibrium systems; learned samplers for rare-event sampling and chemical engineering; and inverse problems approached with diffusion-model-based techniques. The course emphasizes how modern AI methods can be used to model, accelerate, and solve scientific computing problems across domains such as fluid dynamics, climate, protein design, and chemical engineering.
Prerequisites: 10-301 (min C) or 10-315 (min C) or 07-280 (min C) or 10-701 (min C) or 10-601 (min C) or 10-715 (min C)
Offered: Fall 2026
Courseweave needs JavaScript to show sections, workloads and planning tools.