TI TVM User’s Guide - 8.6.x

Texas Instrument’s fork of the Apache Tensor Virtual Machine (TVM) enables support for the TDA4 family of processors. These processors use C7x DSPs and Matrix Multiplication Accelerators (MMA) to accelerate inference-making by machine learning models. For additional information about TDA4x processors and TI’s Edge AI ecosystem, refer to the Edge AI page on ti.com.

_images/TDA4.png

The TI Deep Learning library (TIDL) contains highly optimized implementations of common layers on C7x/MMA. If a model contains layers that are not implemented by TIDL, TVM can be used to run these layers on Arm (Cortex-A) cores or the C7x DSP core.

This user’s guide documents the TI TVM Compiler and its usage.

For offline use, a PDF version of the guide is available here: TI TVM User's Guide.