Installation Guide Overview#

This guide provides complete instructions for installing and deploying TensorRT-RTX on supported platforms. Whether you are setting up TensorRT-RTX for the first time or upgrading an existing installation, this guide walks you through the process.

About TensorRT-RTX#

NVIDIA TensorRT for RTX (TensorRT-RTX) is an inference library that brings TensorRT optimization to NVIDIA RTX GPUs across desktops, laptops, and workstations. It uses a portable ahead-of-time (AOT) engine build plus just-in-time (JIT) compilation on the end-user GPU, so you do not need a separate full engine build for every target device.

For the two-phase compilation pipeline and ecosystem context, refer to the Architecture Overview. For runtime object lifetimes, threading, and memory contracts when embedding TensorRT-RTX, refer to How TensorRT-RTX Works.