getting-started

title	Getting Started
nav_order	1
has_children	true

Getting Started

This section will help you set up, build, and validate TensorCraft-HPC on your system.

In This Section

{: .toc }

• Installation Guide - Complete setup instructions
• Troubleshooting - Common issues and solutions

Quick Start

Choose the build preset that matches your needs:

Development (Recommended)

For CUDA development on a GPU-equipped machine:

cmake --preset dev
cmake --build --preset dev --parallel $(nproc)
ctest --preset dev --output-on-failure

Python Bindings Only

If you mainly need the Python interface:

cmake --preset python-dev
cmake --build --preset python-dev --parallel $(nproc)
python3 -m pip install -e .
python3 -c "import tensorcraft_ops as tc; print(tc.__version__)"

Full Release Build

For benchmarks and complete validation:

cmake --preset release
cmake --build --preset release --parallel $(nproc)
ctest --test-dir build/release --output-on-failure

CPU-Only Validation

To test build infrastructure without CUDA:

cmake --preset cpu-smoke
cmake --install build/cpu-smoke --prefix /tmp/tensorcraft-install

Next Steps

Once you've successfully built the project:

1. Explore Examples → Examples Section
2. Understand Architecture → Architecture Guide
3. Reference API → API Documentation
4. Learn Optimization → Optimization Guides

Prerequisites Overview

Component	Version	Required
CUDA Toolkit	12.0+	Yes (for GPU features)
CMake	3.20+	Yes
C++ Compiler	C++17	Yes
Python	3.8+	No (for bindings)
NVIDIA GPU	Compute 70+	No (for tests/benchmarks)

For detailed prerequisites and troubleshooting, see the Installation Guide.