Numba
JIT compiler for Python
From Wikipedia, the free encyclopedia
Numba is an open-source JIT compiler that translates a subset of Python and NumPy into fast machine code using LLVM, via the llvmlite Python package. It offers a range of options for parallelising Python code for CPUs and GPUs, often with only minor code changes.
| Numba | |
|---|---|
 Numba logo | |
| Original author | Continuum Analytics |
| Developer | Community project |
| Initial release | 15 August 2012 |
| Stable release | |
| Written in | Python, C |
| Operating system | Cross-platform |
| Platform | x86-64, ARM64, POWER |
| Type | Technical computing |
| License | BSD 2-clause |
| Website | numba |
| Repository | |
| Numba CUDA | |
|---|---|
| Developer | NVIDIA |
| Stable release | 0.4.0
/ January 27, 2025[2] |
| Platform | NVIDIA GPU |
| License | BSD 2-clause |
| Website | nvidia |
| Repository | github |
Numba was started by Travis Oliphant in 2012 and has since been under active development with frequent releases. The project is driven by developers at Anaconda, Inc., with support by DARPA, the Gordon and Betty Moore Foundation, Intel, Nvidia and AMD, and a community of contributors on GitHub.
Example
Numba can be used by simply applying the numba.jit decorator to a Python function that does numerical computations:
import numba
import random
@numba.jit
def monte_carlo_pi(n_samples: int) -> float:
"""Monte Carlo"""
acc = 0
for i in range(n_samples):
x = random.random()
y = random.random()
if (x**2 + y**2) < 1.0:
acc += 1
return 4.0 * acc / n_samples
The just-in-time compilation happens transparently when the function is called:
>>> monte_carlo_pi(1000000)
3.14
GPU support
Numba can compile Python functions to GPU code. Initially two backends are available:
- NVIDIA CUDA, see numba
.readthedocs .io /en /stable /cuda /index .html - AMD ROCm HSA, see numba
.pydata .org /numba-doc /dev /roc
Since release 0.56.4,[3] AMD ROCm HSA has been officially moved to unmaintained status and a separate repository stub has been created for it.