In a typical Linux setup, especially on systems with both integrated and dedicated GPUs, it’s common to have applications use the integrated GPU by default to save power. However, there are scenarios where the power of a dedicated NVIDIA GPU is required to run graphic-intensive applications. Fortunately, NVIDIA provides an easy way to offload rendering to the dedicated GPU using the environment variables __NV_PRIME_RENDER_OFFLOAD and __GLX_VENDOR_LIBRARY_NAME. In this post, we will explore what these variables do, and how to use them through a simple script to easily launch applications with the dedicated GPU.
Getting Started with NVIDIA Jetson TX2
I’ve been playing with Jetson for a couple of days. My first objective is to flash it with all required software to be able…