Technology company Nvidia has introduced Project DIGITS, a personal AI supercomputer that fits on a desktop. Designed for AI researchers, data scientists, and students, this system makes it easier to build, test, and use advanced AI models.
“AI will be mainstream in every application for every industry. With Project DIGITS, the Grace Blackwell Superchip comes to millions of developers,” said Jensen Huang, founder and CEO of Nvidia. “Placing an AI supercomputer on the desks of every data scientist, AI researcher and student empowers them to engage and shape the age of AI.”
It is powered by the Nvidia GB10 Grace Blackwell Superchip, which delivers incredible computing power while remaining energy-efficient.
The GB10 Superchip includes an advanced GPU with the latest CUDA and Tensor Cores, paired with a high-performance CPU built using Arm architecture. Together, they offer up to 1 petaflop of AI computing performance. This allows users to work on large AI models, including those with up to 200 billion parameters. By connecting two DIGITS systems, users can even handle models with 405 billion parameters.
Library of AI tools
Project DIGITS includes 128GB of unified memory and up to 4TB of storage, making it capable of running massive AI tasks. It connects easily to cloud or data center systems, letting users seamlessly move their work from local testing to larger-scale deployments.
The system runs on Linux-based Nvidia DGX OS and gives access to a library of AI tools and software, including Nvidia RAPIDS for data science, NeMo for fine-tuning language models, and frameworks like PyTorch. Developers can also use Nvidia’s Blueprints and NIM microservices to create advanced AI applications.
With its user-friendly design and compatibility with Nvidia’s cloud and enterprise systems, Project DIGITS enables experimentation and real-world deployment of AI projects.
Set to be available in May, Project DIGITS will start at $3,000 and will be available through Nvidia and its partners.