Loading...
Loading...
Browse all stories on DeepNewz
VisitWhat will be the primary use case for Nvidia's Project Digits by end of 2025?
AI Research • 25%
Data Science • 25%
Educational Use • 25%
Other • 25%
Market analysis reports or Nvidia's official statements
Nvidia Launches Project Digits with GB10 Grace Blackwell Superchip, Capable of 200 Billion Parameters, for $3,000 in May
Jan 7, 2025, 04:21 AM
At CES 2025, Nvidia Corp. unveiled Project Digits, a personal AI supercomputer designed for AI researchers, data scientists, and students. The device, powered by the new Nvidia GB10 Grace Blackwell Superchip developed in collaboration with MediaTek, offers up to 1 petaflop of AI computing performance at FP4 precision. It is capable of running models with up to 200 billion parameters, and when two units are linked together, they can handle models up to 405 billion parameters. Project Digits, which resembles a Mac Mini in size, will be available starting in May for $3,000. The supercomputer features 128GB of unified memory and up to 4TB of NVMe storage, and runs on Nvidia's Linux-based DGX OS, preconfigured with the Nvidia AI software stack. Nvidia CEO Jensen Huang highlighted that Project Digits aims to bring the power of the Grace Blackwell platform to millions of developers, enabling them to prototype, fine-tune, and run large AI models directly from their desktops. Users can also deploy their models on Nvidia DGX Cloud or other data center infrastructure using the Nvidia AI Enterprise software platform.
View original story
Data Scientists • 25%
Hobbyists • 25%
Students • 25%
AI Researchers • 25%
Education • 25%
Other • 25%
Data Science • 25%
AI Research • 25%
Healthcare Robotics • 25%
Consumer Robotics • 25%
Autonomous Vehicles • 25%
Industrial Robotics • 25%
No significant response • 25%
Launch a similar product • 25%
Price reduction in existing products • 25%
Partnerships for enhanced AI solutions • 25%
Military or defense projects • 25%
AI research and development • 25%
Commercial AI applications • 25%
Other specialized AI applications • 25%
No • 50%
Yes • 50%
Poor (below 5) • 25%
Average (5-6) • 25%
Excellent (9-10) • 25%
Good (7-8) • 25%
No • 50%
Yes • 50%
Healthcare • 25%
Retail • 25%
Manufacturing • 25%
Transportation • 25%
Other • 25%
AMD • 25%
Intel • 25%
Apple • 25%