A maintained, unified, and extended Versatile Tensor Accelerator (VTA) ecosystem.
This repository addresses the limitations of the original VTA project by providing:
- Unified Simulation: A consistent input format (raw binary files) for both functional (C++) and cycle-accurate (CHISEL) simulators.
- Extended Cycle-Accurate Simulation: Enriched cycle-accurate simulation with multiple test cases for different submodules.
- Standalone Compiler: An open-source, TVM-independent compiler for generating VTA binaries.
This project aims to improve the VTA's usability and applicability, particularly in safety-critical systems like aeronautics. The VTA is an open-source hardware accelerator designed to efficiently execute matrix multiplications, a core operation in Convolutional Neural Networks (CNNs).
tutorials/
: Contains a Jupyter tutorial in two parts explaining and illustrating the use of the VTA compiler.src/
: Contains the source code of the project.compiler/
: Contains the VTA compiler that generates the binaries from JSON file.simulators/
: Both functional (C++) and cycle-accurate (CHISEL) simulators.
examples/
: The examples to run.Makefile
: Usemake help
to get the different examples to run.
config/vta_config.json
: The JSON file that defines the VTA hardware parameters.environment_setup\standalone-vta.yml
: The file to setup the conda environment for executing the project.
To get started with this repository, follow these steps:
- Clone the repository:
git clone https://github.com/onera/standalone-vta.git cd standalone_vta
- Run the examples:
It results in two folders:
cd examples make help make matrix_16x16
compiler_output/
andsimulators_output
.