Understanding the Tenstorrent Software Stack
This document provides developers and researchers with an overview of the main components of the Tenstorrent software stack: TT-Metalium™, TT-NN™, and TT-Forge™. You will learn about each tool’s purpose, use cases, and where to find its source code.
TT-Metalium™: Programming Tenstorrent Hardware
TT-Metalium™ is the low-level, open-source software development kit (SDK) that provides developers direct access to Tenstorrent hardware. It is a bare-metal programming environment designed for users who must write custom C++ kernels for machine learning or other high-performance computing workloads.
Use TT-Metalium™ when you require complete control over the hardware to optimize code for performance, explicitly manage memory, or implement novel operations not found in standard libraries. This environment exposes the RISC-V processors, the Network-on-Chip (NoC), and the matrix and vector engines within each Tensix core.
The main project is available in the tt-metal GitHub repository.
TT-NN™: A Python & C++ Neural Network OP library
TT-NN™ is a library of neural network operations that provides a user-friendly interface for running models on Tenstorrent hardware. It is designed to be intuitive for developers familiar with PyTorch.
Use TT-NN™ to run AI models using a familiar, high-level Python API without managing the complexities of the underlying hardware. TT-NN™ builds upon TT-Metalium™ and provides a stable set of pre-packaged, optimized operations. It is also available with a C++ API.
The TT-NN™ library is part of the tt-metal GitHub repository.
TT-Forge™: The MLIR-Based Compiler
TT-Forge™ is Tenstorrent’s Multi-Level Intermediate Representation (MLIR)-based compiler. It bridges high-level machine learning frameworks with the Tenstorrent software stack.
Use TT-Forge™ to compile models from frameworks such as PyTorch, JAX, and TensorFlow for execution on Tenstorrent hardware. It offers an automated, general path to run many types of model architectures without requiring custom kernel development. TT-Forge™ integrates with and lowers to TT-Metalium for hardware execution.
The main project is located in the tt-forge GitHub repository.
Exploring our Developer Hub
To continue learning, visit and explore our Developer Hub. There you will find in-depth articles, and information on our software bounty program. You can also find tutorials and technical overviews on our YouTube channel.
Engaging with our Community
For real-time discussions, community Q&A, and to share your latest projects, we invite you to join our Discord channel. It’s a great place to engage directly with other developers and members of the Tenstorrent community.
As always, if you have any questions, please raise a support request with our team. We’re here to help, and are excited to see what you build.