If you’re looking for your own personal AI supercomputer, Nvidia has you covered.
The chipmaker announced at CES that it will launch a personal AI supercomputer called Project Digits in May. The heart of Project Digits is the new GB10 Grace Blackwell superchip, which packs enough processing power to run sophisticated AI models while being compact enough to fit on a desk and operate from a standard power outlet ( this type of processing power previously required much greater processing power). , more energy-intensive systems). This desktop-sized system can handle AI models with up to 200 billion parameters and has a starting price of $3,000. The product itself looks a lot like a Mac Mini.
“AI will be omnipresent in all applications across all sectors. With Project Digits, the Grace Blackwell Superchip is brought to millions of developers,” Nvidia CEO Jensen Huang said in a press release. “Putting an AI supercomputer on the desk of every data scientist, AI researcher, and student empowers them to engage and shape the AI era. »
Each Project Digits system is equipped with 128 GB of unified and coherent memory (for comparison, a good laptop can have 16 GB or 32 GB of RAM) and up to 4 TB of NVMe storage. For even more demanding applications, two Project Digits systems can be linked together to manage models with up to 405 billion parameters (Meta’s best model, Llama 3.1, has 405 billion parameters).
The GB10 chip delivers up to 1 petaflop of AI performance (meaning it can perform 1 quadrillion AI calculations per second) with FP4 accuracy (which helps speed up calculations by performing approximations ), and the system features Nvidia’s latest generation CUDA cores and fifth generation Tensor Cores, connected via NVLink-C2C to a Grace processor containing 20 Arm cores energy efficient. MediaTek, known for its Arm-based chip designs, collaborated on the development of the GB10 to optimize its power efficiency and performance.
Users will also have access to Nvidia’s library of AI software, including development kits, orchestration tools, and pre-trained models available through the Nvidia NGC catalog. The system runs on the Linux-based Nvidia DGX operating system and supports popular frameworks such as PyTorch, Python and Jupyter notebooks. Developers can refine models using the Nvidia NeMo framework and accelerate data science workflows with Nvidia RAPIDS libraries.
Users can develop and test their AI models locally on Project Digits, then deploy them to cloud services or data center infrastructure using the same Grace Blackwell architecture and Nvidia AI Enterprise software platform.
Nvidia offers a line of similar devices in the same style of accessibility: in December, it announced a $249 version of its Jetson computer for AI applications, targeting hobbyists and startups, called Jetson Orin Nano Super (it manages models with up to 8 billion parameters). .
suisse
