https://github.com/eugenehp/gpu-fft
GPU based FFT written in Rust and CubeCL
https://github.com/eugenehp/gpu-fft
cubecl fft gpu kernel rust
Last synced: 12 months ago
JSON representation
GPU based FFT written in Rust and CubeCL
- Host: GitHub
- URL: https://github.com/eugenehp/gpu-fft
- Owner: eugenehp
- License: mit
- Created: 2025-03-01T04:35:32.000Z (over 1 year ago)
- Default Branch: master
- Last Pushed: 2025-06-11T06:36:05.000Z (about 1 year ago)
- Last Synced: 2025-06-11T07:41:42.028Z (about 1 year ago)
- Topics: cubecl, fft, gpu, kernel, rust
- Language: Rust
- Homepage: https://crates.io/crates/gpu-fft
- Size: 64.5 KB
- Stars: 23
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# GPU-FFT
This project demonstrates the use of the `gpu_fft` library in Rust to perform Fast Fourier Transform (FFT) and Inverse Fast Fourier Transform (IFFT) on a generated sine wave signal. The application calculates the dominant frequencies in the signal and prints them along with their power.
## Table of Contents
- [Features](#features)
- [Requirements](#requirements)
- [Installation](#installation)
- [Usage](#usage)
- [License](#license)
## Features
- Generate a sine wave signal.
- Perform FFT to analyze the frequency components of the signal.
- Calculate the Power Spectral Density (PSD).
- Identify and print the dominant frequencies in the signal.
- Perform IFFT to reconstruct the original signal.
## Roadmap
- [ ] Add twiddles algorithm
## Requirements
- Rust (1.84.1 or later)
- `gpu_fft` crate
## Installation
```base
cargo add gpu_fft -F wgpu
```
## Usage
To run the application, use the following command:
```bash
cargo run --example simple
```
The program will generate a sine wave with a specified frequency and sample rate, perform FFT, and print the dominant frequencies along with their power.
### Example Output
```
====================
Input with frequency - 10 Hz
====================
1000000 [0.0, 0.06279052, 0.12533323, 0.18738133, 0.2486899, 0.309017, 0.36812457, 0.4257793, 0.4817537, 0.5358268]..
====================
FFT 3.7933425s
====================
Frequency: 10.00 Hz, Power: 249999.38
Frequency: 958.99 Hz, Power: 122.58
Frequency: 990.00 Hz, Power: 247388.88
====================
IFFT 4.030771s
====================
```
## License
This project is licensed under the MIT License. See the [LICENSE](/LICENSE) file for details.
## Copyright
©️ 2025, Eugene Hauptmann