Artificial intelligence (AI) is already taking center stage in the new year. At CES 2025, NVIDIA announced Project DIGITS, the world's smallest AI supercomputer powered by the GB10 superchip, which is capable of running 200-billion-parameter models.

Project DIGITS features the new NVIDIA GB10 Grace Blackwell superchip, offering a petaflop of AI computing performance for prototyping, fine-tuning and running large AI models. With Project DIGITS, NVIDIA explains, users can develop and run inference on models using their own desktop system, then seamlessly deploy the models on accelerated cloud or data center infrastructure.

The compact, desktop-sized system should be available in May, starting at a price of $3,000.

 
NVIDIA Project DIGITS
Project DIGITS (Source: NVIDIA)

GB10 Superchip

Consider the following overview and key features of the GB10 Superchip.
 
  • System-on-a-chip (SoC) based on NVIDIA Grace Blackwell architecture.
  • Delivers up to 1 petaflop of AI performance at FP4 precision.
  • NVIDIA Blackwell GPU: Latest-generation CUDA cores and fifth-generation Tensor Cores.
  • NVIDIA Grace CPU: Includes 20 Arm architecture-based power-efficient cores.
  • Connected via NVLink-C2C chip-to-chip interconnect.
  • Developed in collaboration with MediaTek, ensuring power efficiency, performance, and connectivity.

Subscribe
Tag alert: Subscribe to the tag Nvidia and you will receive an e-mail as soon as a new item about it is published on our website!

Project DIGITS

Let's take a look at the capabilities of Project DIGITS.
 
  • Operates using a standard electrical outlet.
  • Includes 128 GB of unified memory and up to 4TB of NVMe storage.
  • Runs large language models up to 200 billion parameters for advanced AI applications.
  • Supports NVIDIA ConnectX networking to link two systems, enabling up to 405-billion-parameter models.

“AI will be mainstream in every application for every industry. With Project DIGITS, the Grace Blackwell Superchip comes to millions of developers,” Jensen Huang, founder and CEO of NVIDIA, said in a release. “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.”

Subscribe
Tag alert: Subscribe to the tag Embedded & AI and you will receive an e-mail as soon as a new item about it is published on our website!