Installation

Add tenferro to your Cargo.toml:

[dependencies]
tenferro = { path = "/path/to/tenferro-rs/tenferro" }

The default feature set enables the cpu-faer backend. This is the easiest starting point for CPU-only use. Once tenferro is published, you can switch the path dependency to a crates.io version.

CPU Backends

Exactly one CPU backend must be enabled:

Feature	Use when
cpu-faer	You want the default pure-Rust CPU backend.
cpu-blas	You want BLAS/LAPACK-backed CPU linear algebra.
src-openblas	You want cpu-blas with OpenBLAS built from source.

To choose BLAS instead of the default backend, disable default features:

[dependencies]
tenferro = { path = "/path/to/tenferro-rs/tenferro", default-features = false, features = ["cpu-blas"] }

For source-built OpenBLAS:

[dependencies]
tenferro = { path = "/path/to/tenferro-rs/tenferro", default-features = false, features = ["src-openblas"] }

Do not enable both cpu-faer and cpu-blas in the same build.

CUDA Feature

CUDA support is enabled with the cuda feature alias:

[dependencies]
tenferro = { path = "/path/to/tenferro-rs/tenferro", features = ["cuda"] }

CUDA support targets NVIDIA CUDA through the CubeCL backend. See Devices and GPU for the transfer model and current coverage.

Common CUDA runtime environment variables:

CUBECL_DEBUG_LOG=0
CUDA_PATH=/usr/local/cuda-12.0
LD_LIBRARY_PATH=/usr/local/cuda-12.0/lib64:$LD_LIBRARY_PATH

If CUDA libraries are installed outside the default search paths, set TENFERRO_CUTENSOR_PATH, TENFERRO_CUSOLVER_PATH, or TENFERRO_CUBLAS_PATH to the concrete library path.