Scientific computing with Python: Tools for the solution of continuous problems
summary
Shows how Python solves real large PDEs not just theory: matrices and vectors from thousands up to billions of entries, finite‑difference prototypes sped to C speeds with NumPy and GPU CUDA kernels, loops accelerated by Numba/JIT or Hope to within a few× of C, and finite‑element forms compiled to C (FEniCS‑style form compiler) to assemble matrices efficiently. Demonstrates working codebases, performance tradeoffs, and when each tool and discretization actually succeeds.
AI-written summary · about
chapters (10)
- 0:00Scientific computing intro
- 1:58Micromagnetism and partial differential equations
- 8:24Continuous modeling and discretization
- 11:36Finite difference for Poisson
- 17:01Performance with NumPy
- 22:52Optimizing loops with Numba/Cython
- 27:21Integrating C code
- 29:33Finite element method
- 34:36Python tools for scientific computing
- 35:40Q&A
related talks
01h 31mModern scientific computing and big data analytics in Python
33mThe Infiniteness of Finite Elements Explored through FEniCS
55mAccelerated Python: The Community and Ecosystem
35mPython at the Speed of Light: A Complete Tour of the CUDA Ecosystem
44mPython's Bright Future in Science
27mSpeed up Python with hardware accelerators
31m"Optimizing Scientific Python with C++, CUDA, and Serverless Compute"
02h 55mGPU Accelerated Python w: CUDA