Nvidia’s Digits is a small AI computer for your desk

With all eyes on Nvidia’s new cards (and for good reason – these things are very powerful), the company released something that could end up being the biggest news: An AI personal computer called Project Digits (we’ll just call it Digits from now on because that’s cool).

Digits is a Mac mini-like personal computer that should fit on every desk. You plug it into a keyboard and monitor, plug it into power, and you’re ready to go.

But it’s what’s inside that makes it so special. The digits are powered by Nvidia’s new GB10 Grace Blackwell Superchip, which delivers one petaflop of AI performance, and Nvidia teamed up with MediaTek to make the chip more energy-efficient, meaning that running it requires the kind of power you’d get on a regular one. electricity.

The GB10 is paired with 128GB of RAM and up to 4TB of NVMe storage. All of this will enable developers to use large-scale linguistic models (LLMs) with up to 200-billion-parameters. Using Nvidia ConnectX technology, they will also be able to connect two Digits computers to run up to 405-billion-parameter models.

Most of these things probably make little sense to people who are not in the business of developing an AI application. But for example, ChatGPT 3.5 had 175 billion parameters (ChatGPT 4 is much bigger but we don’t know the exact numbers), while Meta’s most powerful LLM, Llama 3, has 405 billion parameters). This means you can run the most powerful LLM at home instead of relying on cloud infrastructure.

“Putting an AI supercomputer on the desks of every data scientist, AI researcher and student empowers them to engage and shape the age of AI,” said Jensen Huang, founder and CEO of Nvidia, in a written statement.

Developers using Digits will have access to Nvidia’s library of AI software including development tools, orchestration tools, frameworks and models available in the Nvidia NGC catalog and the Nvidia Developer Portal. They will also have access to the Nvidia NeMo framework, as well as the Nvidia RAPIDS libraries.

The big news here, perhaps, is the price. Digits will be available in May, starting at $3,000. This kind of money should make it available to a large number of small companies and researchers, who will use it to create and test AI applications.

If this still sounds too expensive, you can try Nvidia’s Jetson. It’s a $249 AI home computer launched last December, and it can handle up to 8 billion parameters.

Mashable is in the world live at CES 2025! We’re covering all the wildest and most important events this week, so please keep checking back You are with us. Want to submit a product you represent for our teams to consider as we identify the Best of CES? Here are more details on how to do it.