NVIDIA has unveiled Project G-Assist, an AI model that runs on the GPU RTX without the need for an NPU. This tiny hardware assistant is designed to help adjust various settings on the hardware front.
Project G-Assist is currently in experimental status as a feature of the NVIDIA App. Users can command it to adjust fan speed, change keyboard backlight colors, tweak in-game graphics settings, provide machine specs information, and more. It supports commands via text chat input and voice control.
G-Assist also controls hardware from certain partner brands like Logitech G, Corsair, MSI, and Nanoleaf by allowing connectivity through plugins for integration with other systems.
Behind the scenes, it operates as a Small Language Model (SLM) focusing on limited functionality, unlike the familiar Large Language Model (LLM) models.
This feature is compatible with GeForce RTX 30 series and above, requiring a minimum of 12GB of RAM.
Source: NVIDIA, NVIDIA
TLDR: NVIDIA introduces Project G-Assist, a small AI model that assists with hardware settings and controls, running on GeForce RTX GPUs with 12GB RAM or more.
Leave a Comment