Conquer3D

Installation

1. Install from PyPI

pip install -U conquer3d

2. Install via Docker (Recommended for complete 3D & CUDA setup)

You can directly pull and run the pre-built Docker image with full GPU and CUDA toolchain support:

docker pull kohido/conquer3d:latest
docker run --rm --gpus all -it kohido/conquer3d:latest bash

Or build the image locally from source:

docker build -t conquer3d:latest .
docker run --rm --gpus all -it conquer3d:latest bash

3. Build from Source

To build from source, ensure you have a compatible CUDA toolchain (e.g., CUDA Toolkit 12.8) and PyTorch installed:

# Optional: Create a dedicated Conda environment with modern C++ compilers and CUDA
conda create -c conda-forge -n geocutool python=3.10 gxx_linux-64=13 gcc_linux-64=13 -y
conda activate geocutool
conda install -c conda-forge sparsehash -y
conda install nvidia::cuda-toolkit==12.8.2 -y

# Install PyTorch and binding generators
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu128
pip install pybind11-stubgen

Install directly from GitHub without build isolation:

pip install git+https://github.com/KhoiDOO/geocutool.git --no-build-isolation

git clone https://github.com/KhoiDOO/geocutool.git
cd geocutool
pip install -e . --no-build-isolation

Features

Isosurface Extraction: Differentiable Marching Cubes, Marching Cubes Asymptotic (MCA), Marching Tetrahedra (MT), and Dual Contouring (DC) with GPU Jacobi SVD Quadratic Error Function (QEF) solver for sharp feature preservation and pure quad extraction (quad_split=False).
Spatial Acceleration Structures: GPU Bounding Volume Hierarchies (Mesh BVH, GS BVH, PGS BVH), KD-Trees, Morton Z-Curve sorting, and 3D Sparse Voxel Grids.
Differentiable Ray Tracing & Radiance Fields: Fast 3D Gaussian Splatting (3DGS) and Periodic Gaussian Splatting (PGS) query kernels.
Geometric Metrics & Integrals: Exact one-sided & symmetric Chamfer / Hausdorff distances and single-view volume integrals.

Acknowledgements & References

For further theoretical background, GPU collision detection guides, and related open-source projects, please refer to:

Research Papers: Key computational geometry, differential topology, and acceleration structure literature.
Blog Posts: Articles and guides on NVIDIA GPU spatial traversal and parallel construction.
Related Repositories: Open-source libraries and frameworks supporting geometric deep learning and processing.
Books: Core theoretical books.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

conquer3d-0.5.7.tar.gz (4.3 MB view details)

Uploaded Aug 18, 2026 Source

File details

Details for the file conquer3d-0.5.7.tar.gz.

File metadata

Download URL: conquer3d-0.5.7.tar.gz
Upload date: Aug 18, 2026
Size: 4.3 MB
Tags: Source
Uploaded using Trusted Publishing? No
Uploaded via: twine/7.0.0 CPython/3.13.9

File hashes

Hashes for conquer3d-0.5.7.tar.gz
Algorithm	Hash digest
SHA256	`1a108ac4f10af3c9fa5e2bc30e6bea27390630223cfb859bb13a8678fa57fef8`
MD5	`b3cd4bc7eb3a8ff31f9170b4250b1e2c`
BLAKE2b-256	`e2557289bdf3624aa5dcaf6a152c0b0c2c4c0df064e9553513d09f7f053c2420`