A CES 2025 in Las VegasNvidia unveiling Project Digits, a “personal AI supercomputer” that provides access to the company's Grace Blackwell hardware platform in a compact form factor.
“[Project Digits] The whole Nvidia AI stack runs — all the Nvidia software runs on it,” Nvidia CEO Jensen Huang said on stage during a press conference Monday. “It's a cloud computing platform that sits on your desk … it's even a workstation, if you want it to be.”
Designed for AI researchers, data scientists and students, Project Digits packs Nvidia's new GB10 Grace Blackwell Superchip, which delivers up to petaflops of computing performance for prototyping, fine-tuning and running AI models.
Nvidia claims that a single project can model up to 200 billion parameters in digit units. Parameters roughly correspond to a model's problem-solving ability, and models with more parameters generally perform better than models with fewer parameters.
The GB10, which is developed in partnership with MediaTek, features an Nvidia Blackwell GPU paired with a 20-core Nvidia Grace CPU. Inside the Project Digits enclosure, the chips are connected to a pool of up to 128GB of memory and up to 4TB of flash storage.
Nvidia says that two Project Digit machines can be linked together to run a 405-billion-parameter model, should a task require it. Project Digits can provide a standalone experience, as mentioned earlier, or connect to a basic Windows or Mac PC.
But it's not cheap. Project Digit machines running Nvidia's Linux-based DGX OS will be available in May from “top partners” for $3,000, the company said.
So it won't be just anyone who can afford a Project Digit unit of their own. Huang thinks that there is a market, however.
“With Project Digit, the Grace Blackwell Superchip comes to millions of developers,” he said in a statement. “Putting an AI supercomputer on the desk of every data scientist, AI researcher and student empowers them to engage and shape the age of AI.”