NVIDIA announces new chips and tools for on-device AI

January 9, 2024

NVIDIA used the opening day of CES to announce a range of new GPUs, graphics cards, and developer tools aimed at building and running AI models and applications on local devices.

Most of the attention the company got last year was focused on the demand for its GPUs that power large data centers. With this announcement, NVIDIA is going back to its roots with a focus on powering generative AI on laptops and PCs.

Its GeForce RTX SUPER desktop GPUs will power a range of new laptops from manufacturers including Acer, ASUS, Dell, HP, Lenovo, MSI, Razer, and Samsung.

Tensor Cores in these GPUs offer significant AI performance improvements which NVIDIA says range from “20x-60x compared with using neural processing units.” NVIDIA says the first of these laptops will ship later this month.

For desktop PC users NVIDIA announced its new GeForce RTX 40 SUPER Series graphics cards. The RTX 4060 Super, RTX 4070 Ti Super, and RTX 4080 Super range in price between $599 and $999. These cards all fly under the radar of the US export controls restricting AI tech supplied to China.

The GeForce RTX 4080 SUPER generates AI video 1.5x faster, and images 1.7x faster than its predecessor the GeForce RTX 3080 Ti GPU.

Having hardware that is capable of running LLMs locally has huge benefits.

  • Cost: No cloud-hosted API or infrastructure costs for LLM inference.
  • Always-on: No internet connectivity is required.
  • Performance: Low latency or even real-time inference.
  • Data privacy: Data can always stay on the device instead of passing through cloud servers.

New AI developer tools

Jensen Huang, founder and CEO of NVIDIA said, “Generative AI is the single most significant platform transition in computing history and will transform every industry, including gaming. With over 100 million RTX AI PCs and workstations, NVIDIA is a massive installed base for developers and gamers to enjoy the magic of generative AI.”

In other words, there are plenty of users with powerful hardware just waiting for AI-powered games and applications.

NVIDIA’s AI Workbench will be available in beta later this month. It offers developers a simplified interface to develop projects on Hugging Face, GitHub, and NVIDIA NGC.

It allows developers to develop, run, and migrate their projects back and forth between data centers and local machines.

AI is already disrupting how computer games are developed. NVIDIA RTX Remix will breathe new life into your favorite classic games.

The tool captures scenes and elements in an older game and upscales the resolution up to 4K. It also improves textures and details and allows modification of elements in the game.

This demo shows how amazing some of these older games can look after a little generative AI magic.

Staying with gaming, NVIDIA also announced NVIDIA Avatar Cloud Engine (ACE). This tool enables game developers to use generative AI to create digital avatars that run in the cloud or locally on the PC. These avatars work seamlessly with third-party applications like Unreal Engine 5.

Chat with RTX will also be available later this month. This uses retrieval-augmented generation (RAG) to connect an LLM running on a PC with a user’s data. Developers could use this to create applications like personal assistants where the locally run LLM can answer questions specific to user data.

The CES tech event has only just started so we can expect more exciting announcements from NVIDIA and other companies this week.

Join The Future


SUBSCRIBE TODAY

Clear, concise, comprehensive. Get a grip on AI developments with DailyAI

Eugene van der Watt

Eugene comes from an electronic engineering background and loves all things tech. When he takes a break from consuming AI news you'll find him at the snooker table.

×

FREE PDF EXCLUSIVE
Stay Ahead with DailyAI

Sign up for our weekly newsletter and receive exclusive access to DailyAI's Latest eBook: 'Mastering AI Tools: Your 2024 Guide to Enhanced Productivity'.

*By subscribing to our newsletter you accept our Privacy Policy and our Terms and Conditions