All Tools
C
Dev ToolsFreeOpen Source
CONDA
Cross-platform package and environment management
BSD-3-Clause
ABOUT
Python environments on HPC clusters and data science workstations need non-Python dependencies (CUDA, BLAS, OpenMP, HDF5) that pip cannot install, and system package managers (apt, brew) install incompatible versions. Conda provides language-agnostic package management with pre-built binaries for Linux, macOS, and Windows. It resolves and installs Python packages AND their native library dependencies (cuDNN, MKL, libgcc) in isolated environments. The result is reproducible scientific computing stacks that include C/C++/Fortran libraries, GPU toolkits, and Python packages in one atomic operation.
INSTALL
brew install --cask minicondaINTEGRATION GUIDE
1. Install and isolate ML environments with GPU dependencies (CUDA, cuDNN, NCCL) alongside PyTorch or TensorFlow in one command
2. Manage Python versions and package sets for Jupyter notebook environments on shared HPC or academic compute clusters
3. Create reproducible conda environment YAML files for research projects that pin both Python and C library versions
4. Distribute scientific Python software that depends on native C/C++/Fortran libraries (GDAL, netCDF, VTK) across platforms
5. Build and deploy containerized data science pipelines using conda-pack for portable environment snapshots
TAGS
pythonpackage-managerenvironment-managementdata-sciencecross-platformdeveloper-tools