NVIDIA Project Gr0ok: Your Personal AI Supercomputer

Nvidia Unveils Project Digits: A Personal AI Supercomputer

During CES 2025 in Las Vegas, Nvidia introduced Project Digits, a uniquely compact “personal AI supercomputer.” This device grants access to the company’s powerful Grace Blackwell hardware platform.

Key Features and Capabilities

According to Nvidia CEO Jensen Huang, “[Project Digits] operates the complete Nvidia AI stack – encompassing all of Nvidia’s software.” He described it as a desktop-based cloud computing platform, and also a fully functional workstation.

Project Digits is specifically engineered for AI researchers, data scientists, and students. It integrates Nvidia’s latest GB10 Grace Blackwell Superchip, capable of delivering up to a petaflop of computational performance.

Model Size and Performance

This performance enables prototyping, fine-tuning, and the execution of complex AI models. Nvidia asserts that a single Project Digits unit can effectively manage models containing up to 200 billion parameters.

The number of parameters generally indicates a model’s capacity for problem-solving; larger models typically outperform those with fewer parameters. Therefore, this capability is significant.

Hardware Specifications

The GB10, developed in collaboration with MediaTek, combines an Nvidia Blackwell GPU with a 20-core Nvidia Grace CPU. Within the Project Digits enclosure, these chips are connected to 128GB of memory and up to 4TB of flash storage.

Scalability and Connectivity

For demanding tasks, two Project Digits machines can be interconnected to support models with up to 405 billion parameters. The system can function independently or connect to an existing Windows or Mac PC.

Pricing and Availability

Project Digits operates on Nvidia’s Linux-based DGX OS and will be available starting in May through select “top partners.” The price point is $3,000 per unit.

Market Impact

While the cost is substantial, Huang believes a considerable market exists. He stated that Project Digits brings the Grace Blackwell Superchip within reach of a wider audience.

“Empowering every data scientist, AI researcher, and student with an AI supercomputer on their desk will enable them to contribute to and shape the future of AI,” Huang explained.