NVIDIA has just released the first public version of the Chat with RTX demo. But no, you can’t chat with your graphics card and ask, “how are you?”. For that, you need the TechPowerUp GPU-Z program. Chat with RTX is something entirely different. Imagine a fully localized AI chat that runs entirely on your computer, powered by the powerful cores in your GeForce RTX graphics card, and doesn’t send any queries to cloud chat servers. That’s what Chat with RTX is. NVIDIA is working on making it an alternative to ChatGPT, which stores all its knowledge locally on your computer, utilizing the GeForce RTX GPU as its brain.
By the end of 2024, the promised “AI PC year” as predicted by industry leaders Microsoft and Intel, NVIDIA already has a remarkable advantage for six years in AI acceleration. The company introduced locally accelerated AI for real-time ray tracing technology with RTX. As part of this innovation, its 2017 GeForce RTX graphics cards were equipped with Tensor cores. These components significantly enhanced the building and training of deep neural network (DNN) AI compared to using CUDA cores alone. This progressive technology was a significant leap forward in terms of performance, increasing the capabilities of graphics cards for AI-driven tasks. In addition to noise reduction, NVIDIA utilizes AI acceleration to power DLSS performance enhancement features. Can’t fully utilize the game? Enable DLSS and choose one of the available settings until the game runs smoothly at your desired settings.
In our recent conversations with NVIDIA, the company clearly stated that it is not too impressed with the latest processors from Intel and AMD, which introduce neural processing units (NPU); they offer performance ranging from 10 to 16 TOPS for the NPU alone and no more than 40 TOPS for the whole system (NPU + CPU + iGPU). NVIDIA GeForce RTX models, thanks to their Tensor cores, offer 20x to 100x (!) higher performance through massive AI acceleration in their GPU architecture. While CPU-based NPU units aim to handle simple text and image-based AI tasks, NVIDIA employs AI on an entirely different level, even today. Just think about generating every other frame in DLSS 3 Frame Generation technology or reducing noise in a 4K scene in real-time gameplay at over 60 frames per second, depending on the resolution. In other words, GeForce RTX graphics cards have enormous hardware AI acceleration capabilities that remain unused when you’re not gaming. That’s why NVIDIA has decided to show gamers that it is possible to run fully localized AI tools utilizing this hardware acceleration. The company is just getting started, and one of its first projects is Chat with RTX, which we are presenting in a preview form today. With millions of gamers using GeForce RTX graphics cards, NVIDIA’s huge user base means that we can expect the company to take a more active role in the PC AI ecosystem, providing additional AI experiences and tools for computers equipped with GeForce RTX graphics cards.
As previously mentioned, Chat with RTX is a text-based generative AI platform, similar to ChatGPT or Copilot, but one that doesn’t send any data to the cloud or use internet data. The data set is what you provide. You even have the option to choose between the Llama2 and Mistral AI models. For the technical Chat with RTX demo, NVIDIA provides Llama2 and Mistral along with their data models, which are updated until mid-2022.
In this article, we will test Chat with RTX to show you the capabilities it has in delivering a powerful, fully offline AI chat for gamers.
NVIDIA has released a public version of the Chat with RTX demo, aiming to offer localized generative AI chat for gamers. Chat with RTX operates without a cloud connection and utilizes the powerful GeForce RTX GPU cores for acceleration. NVIDIA introduced locally accelerated AI for real-time ray tracing technology, using Tensor cores in GeForce RTX graphics cards, which significantly increased performance in AI tasks. Chat with RTX is one of the first projects within this solution. With a large user base using GeForce RTX graphics cards, NVIDIA plans to expand the PC AI ecosystem by providing additional AI experiences and tools for computers with these cards.
Chat with RTX is a text-based generative AI platform, similar to ChatGPT and Copilot, but it does not rely on the cloud or internet-based data. Users provide the data, which serves as the dataset. In the technical Chat with RTX demo, NVIDIA offers two AI models: Llama2 and Mistral, along with their updated data until mid-2022.
We will test the Chat with RTX demo to showcase the powerful, fully offline AI chat it can deliver for gamers.
FAQ:
1. What is Chat with RTX?
Chat with RTX is a text-based generative AI platform that offers localized AI chat for gamers.
2. How does Chat with RTX utilize GeForce RTX GPU cores?
Chat with RTX utilizes the powerful GeForce RTX GPU cores for AI chat acceleration.
3. Does Chat with RTX rely on the cloud or internet data?
No, Chat with RTX operates fully offline and does not rely on the cloud or internet data.
4. What AI models are available in Chat with RTX?
The available AI models in Chat with RTX are Llama2 and Mistral.
5. How can I provide data for Chat with RTX?
As a user, you can provide data that will be used by Chat with RTX.
6. What capabilities does Chat with RTX offer for gamers?
Chat with RTX offers powerful, fully offline AI chats for gamers.
Related links:
– NVIDIA Homepage
The source of the article is from the blog radardovalemg.com