Nvidia CEO Jensen Huang stated onstage during a press conference on Monday
“Project Digits runs the entire Nvidia AI stack — all of Nvidia software runs on this,”. “It’s a cloud computing platform that sits on your desk … It’s even a workstation if you like it to be.”
Nvidia’s new GB10 Grace Blackwell Superchip, which offers up to a petaflop of computational power for developing, optimizing, and executing AI models, is included into Project Digits, a platform intended for AI researchers, data scientists, and students.
According to Nvidia, one Project Digits unit can execute models with up to 200 billion parameters. A model’s ability to solve problems is largely reflected in its parameters, and models with more parameters typically outperform those with fewer.
The GB10, created in collaboration with MediaTek, has a 20-core Nvidia Grace CPU coupled to an Nvidia Blackwell GPU. The chips are connected to up to 4TB of flash storage and a 128GB pool of memory inside the Project Digits shell.
According to Nvidia, if a work requires it, two Project Digits computers can be connected to execute models with up to 405 billion parameters. As previously mentioned, Project Digits can provide a stand-alone experience or link to a core Windows or Mac computer.
It’s not inexpensive, though. According to the business, Project Digits computers, which run Nvidia’s Linux-based DGX OS, will be offered for $3,000 by “top partners” beginning in May.
Therefore, not everyone will be able to purchase a Project Digits device. But Huang believes there is a market. He said in a statement
“With Project Digits, the Grace Blackwell Superchip comes to millions of developers,”. “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.”
Read Also:
