Not a long time ago, in a galaxy not very far away, an astronomer and a computer scientist walk into
summary
A C++ galaxy-kinematic fitter is ported to Python with most model code in pure Python and critical inner loops as either C++ extensions or Numba/CuPy GPU paths. Metadata uses JSON, astropy and scikit-image handle IO/preprocessing, and multiple optimizers and MCMC libraries plug in. Result: complex high‑resolution, irregular galaxies that previously took hours or days now fit in 1–10 minute runs, showing the approach works and scales across CPU multithreading, OpenMP, and CUDA.
AI-written summary · about
chapters (14)
- 0:00Intro
- 0:15Who are you
- 0:55Why Im here
- 1:17Galaxy Basics
- 2:59Galaxy Kinematics
- 5:21Galaxy Omatic Modelling
- 6:30Challenges
- 9:44Software
- 10:33Languages
- 11:30Python
- 12:37Problems
- 15:55Writing in Python
- 18:43Models
- 23:48Optimizer
related talks
49mPython for High Performance and Scientific Computing
22mDiscovering the mysteries of the universe with Python
31m"Optimizing Scientific Python with C++, CUDA, and Serverless Compute"
35mPython at the Speed of Light: A Complete Tour of the CUDA Ecosystem
30mFrom Matlab to Lambda: Transforming Structural Engineering Research Tools
35mCUDA in Python- A New Era for GPU Acceleration
39mAccelerated Data Science Libraries
30mCuPy: A NumPy-compatible Library for GPU