NVIDIA Brings AI Power to Your Desktop with Project DIGITS
NVIDIA has introduced Project DIGITS, a personal AI supercomputer designed for researchers, developers, and students to work with large AI models directly from their desktops. This compact yet powerful system is built around the NVIDIA GB10 Grace Blackwell Superchip, offering up to 1 petaflop of AI performance and capable of handling models with up to 200 billion parameters.
Powerful Hardware for AI Development
At the core of Project DIGITS is the GB10 Superchip, which integrates an NVIDIA Blackwell GPU with CUDA cores and fifth-generation Tensor Cores, connected via NVLink-C2C to a Grace CPU with 20 Arm-based cores. This combination ensures high efficiency and computational power, making it ideal for AI training and inference tasks.
Each DIGITS unit comes with
- 128GB of unified memory and up to 4TB of NVMe storage, ensuring smooth performance for demanding applications.
- Two units can be linked via NVIDIA ConnectX networking, expanding capabilities to 405 billion parameters.
AI Software and Developer Tools
Project DIGITS runs on Linux-based NVIDIA DGX OS and provides access to NVIDIA’s extensive AI software ecosystem, including tools for AI development, model orchestration, and pre-trained frameworks from the NVIDIA NGC catalog. It supports popular AI frameworks like PyTorch, Python, and Jupyter notebooks, making it a versatile tool for AI researchers and developers.
Affordability and Accessibility
Set to launch in May 2025, Project DIGITS starts at $3,000, making high-performance AI computing more accessible to individuals and small teams. With this release, NVIDIA aims to bring supercomputing capabilities to every AI developer’s desk, accelerating innovation in AI research and applications.
