https://github.com/0xhilsa/pynum
a small python library for 1D and 2D arrays with GPU support
https://github.com/0xhilsa/pynum
array c cuda nvcc python3
Last synced: 2 months ago
JSON representation
a small python library for 1D and 2D arrays with GPU support
- Host: GitHub
- URL: https://github.com/0xhilsa/pynum
- Owner: 0xhilSa
- License: mit
- Created: 2024-12-19T18:15:19.000Z (over 1 year ago)
- Default Branch: master
- Last Pushed: 2025-03-29T19:55:37.000Z (about 1 year ago)
- Last Synced: 2025-10-21T19:25:00.752Z (8 months ago)
- Topics: array, c, cuda, nvcc, python3
- Language: C
- Homepage:
- Size: 399 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Changelog: changelogs.md
- License: LICENSE
Awesome Lists containing this project
README
a small python library for 1D and 2D arrays with GPU support
## 🚀 Accelerators
- CUDA
## 📌 Prerequisites
- GCC Compiler
- NVCC Compiler [NVIDIA TOOLKIT](https://developer.nvidia.com/cuda-downloads)
## ⚡ Installation
```bash
git clone "git@github.com:0xhilSa/pynum.git" ~/pynum
cd pynum && rm -rf .git*
bash install.sh -y
```
## ✅ ToDo
- [X] Implement the `__setitem__` method
- [ ] Develop the `astype` method
- [X] To solve (core dumped) issue while adding 2 vectors of dtype complex128
- [ ] Add arithmetic and logical operations for the `Vector` class
- [ ] Ensure compatibility with Windows OS
## 🔥 Future Enhancements
- Introduction to Matrix dtype
- Implement an LLVM-based accelerator for CPU execution
> [!IMPORTANT]
> Arithmetic ops on vectors can only be performed if they have the same data type
> - (short - short) ✔️
> - (short - int) ❌
## 📜 LICENSE
[MIT](./LICENSE)