Skip to content
Examples using Numba.
Jupyter Notebook Python
Branch: master
Clone or download
Pull request Compare This branch is 8 commits ahead of harrism:master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.gitignore
README.md
cpu_mandelbrot.py
gpu_cuda_mandelbrot.py
gpu_numba_mandelbrot.py
mandelbrot_numba.ipynb

README.md

numba_examples

No code changes, except for making the print statements Python 3 compatible. The .py files are added to allow running of the programs without Jupyter in terminal mode. If you don't want the image to be displayed, simple comment out the show() line.

An observation: The speed up between the numba and the cuda version of the code did show up on my systems (Xeon E3 skylake, and AMD broadwell 8 cores each with Nvidia Quadro M4000 GPU) only in the Jupyter notebook. Running the code standalone in terminal had the same runtime between the two versions. Note: The figures were optained on a Linux Ubuntu 18.10 and Windows 10 systems with Cuda versions 10.1 and 9.0 respectively.

You can’t perform that action at this time.