Nvidia has released an early version of its new AI chatbot called “Chat with RTX,” available as a free demo download for Windows users with a compatible RTX GPU. The chatbot lets users personalize an AI assistant by connecting local files on their PC as a dataset to generate relevant answers based on their data.Â
Chat with RTX leverages Nvidia’s TensorRT software and RTX GPU acceleration to enable generative AI capabilities locally on a user’s Windows PC. It uses a retrieval-augmented generation (RAG) technique to search large datasets and provide relevant responses.
Nvidia’s tool accommodates various document formats, such as .txt, .pdf, .doc, and others, aiming to uphold user privacy through local data processing. This approach enables users to leverage AI capabilities while safeguarding their personal information, as emphasized by Nvidia.
To run Chat with RTX, users need a GeForce RTX 30 or 40-series GPU with at least 8GB of VRAM. Nvidia has invited developers to explore speeding up large language models using RTX GPUs on GitHub.
The release comes as Nvidia’s value has tripled in recent months. Its graphics chips remain in high demand globally. However, despite U.S. export bans, the company has faced controversy over sales to Chinese entities.
Also Read: Nvidia, NSF Launch $30M NAIRR to Democratize AI Research
