Powered by new GB10 superchip, NVIDIA Project DIGITS debuts as world’s smallest AI supercomputer capable of running 200B parameter models
CES—NVIDIA today announced NVIDIA® Project DIGITS, a personal AI supercomputer that gives AI researchers, data scientists, and students around the world access to the power of the NVIDIA Grace Blackwell platform.
Project DIGITS is powered by the new NVIDIA GB10 Grace Blackwell superchip, delivering petaflops of AI computing performance for prototyping, fine-tuning, and running AI models at scale.
Project DIGITS allows users to develop and run model inference using their own desktop systems and seamlessly deploy models to accelerated cloud or data center infrastructure.
“AI will become mainstream in every application in every industry, and Project DIGITS will bring the Grace Blackwell superchip to millions of developers,” said Jensen Huang, founder and CEO of NVIDIA. “Putting an AI supercomputer on the desk of every data scientist, AI researcher, and student will empower them to tackle and shape the age of AI.”
GB10 superchip delivers petaflops of power-efficient AI performance
The GB10 Superchip is a system-on-chip (SoC) based on the NVIDIA Grace Blackwell architecture that delivers up to 1 petaflops of AI performance with FP4 accuracy.
GB10 is powered by the NVIDIA Blackwell GPU with the latest generation CUDA® cores and 5th generation Tensor Cores, with 20 power-efficient cores integrated through the NVLink®-C2C chip-to-chip interconnect. connected to a high-performance NVIDIA Grace™ CPU. Adopts Arm architecture. MediaTek, a market leader in Arm-based SoC designs, collaborated on the design of GB10, contributing to best-in-class power efficiency, performance, and connectivity.
The GB10 superchip allows Project DIGITS to deliver powerful performance using only standard electrical outlets. Each Project DIGITS features 128 GB of integrated coherent memory and up to 4 TB of NVMe storage. Supercomputers allow developers to run large language models with up to 200 billion parameters to accelerate AI innovation. Additionally, with NVIDIA ConnectX® networking, you can link two Project DIGITS AI supercomputers to run up to 405 billion parameter models.
Grace Blackwell AI supercomputing within reach
The Grace Blackwell architecture allows companies and researchers to prototype, fine-tune, and test models on local Project DIGITS systems running the Linux-based NVIDIA DGX OS, then deploy them to the NVIDIA DGX Cloud™, accelerated It can be seamlessly deployed on any cloud instance or data center infrastructure.
This allows developers to prototype AI with Project DIGITS and scale it on cloud or data center infrastructure using the same Grace Blackwell architecture and NVIDIA AI Enterprise software platform.
Project DIGITS users have access to an extensive library of NVIDIA AI software for experimentation and prototyping, including software development kits, orchestration tools, frameworks, and models available in the NVIDIA NGC Catalog and NVIDIA Developer Portal. Developers can fine-tune models with the NVIDIA NeMo™ framework, accelerate data science with the NVIDIA RAPIDS™ library, and run popular frameworks such as PyTorch, Python, and Jupyter notebooks.
To build agent AI applications, users can also take advantage of NVIDIA Blueprints and NVIDIA NIM™ microservices, which are available for research, development, and testing through the NVIDIA Developer Program. When you’re ready to move your AI applications from lab to production, NVIDIA AI Enterprise licenses provide enterprise-grade security, support, and product releases of NVIDIA AI software.
availability
Project DIGITS will be available in May from NVIDIA and top partners starting at $3,000. Sign up for notifications now.
