At CES 2025 in Las Vegas, Nvidia announced 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 entire Nvidia AI stack is running, and all Nvidia software runs on top of this,” Nvidia CEO Jensen Huang said on stage during a press conference Monday. “It's a cloud computing platform that sits on your desk, or even your workstation if you want to.”
Designed for AI researchers, data scientists, and students, Project Digits is powered by Nvidia's new GB10 Grace Blackwell superchip, delivering up to petaflops of computing performance for prototyping, fine-tuning, and running AI models. It will come true.
Nvidia claims that a single Project Digits unit can run models up to 200 billion parameters in size. Parameters roughly correspond to a model's problem-solving skills, and models with more parameters generally perform better than models with fewer parameters.
Developed in partnership with MediaTek, the GB10 features an Nvidia Blackwell GPU connected to a 20-core Nvidia Grace CPU. Inside the Project Digits enclosure, the chip is connected to a 128 GB memory pool and up to 4 TB of flash storage.
Image credit: Nvidia
Nvidia says that two Project Digits machines can be linked to run up to 405 billion parameter models, depending on the job's needs. Project Digits can offer a standalone experience as mentioned above, or it can be connected to your primary Windows or Mac PC.
But it's not cheap. Project Digits machines running Nvidia's Linux-based DGX OS will be available from “top partners” starting in May for $3,000, the company said.
Therefore, not everyone can purchase a unit of Project Digits for themselves. But Hwang believes the market exists.
“Project Digits brings the Grace Blackwell superchip to millions of developers,” he said in a statement. “By putting an AI supercomputer on the desk of every data scientist, AI researcher, and student, we will be able to tackle and shape the age of AI.”