May 2019
Intermediate to advanced
452 pages
12h 16m
English
Programming in Python with Numba, or Accelerate, does not mandate the use of CUDA-C or OpenCL-C syntax within your Python code, though knowledge of those platforms is definitely an added advantage. With Numba on Anaconda, you can program across heterogeneous compute devices, including NVIDIA and AMD GPUs.
It comes with both CUDA (cudatoolkit Python package) and ROCm (rocmtools Python package) support, though Numba rocmtools is currently in an experimental stage. Numba provides the following features:
Read now
Unlock full access