NVIDIA Unveils DGX Personal AI Supercomputers
NVIDIA has introduced the NVIDIA DGX personal AI supercomputers powered by the NVIDIA Grace Blackwell platform. The DGX Spark, previously known as Project DIGITS, and the DGX Station are high-performance desktop supercomputers that allow AI developers, researchers, data scientists, and students to work on large models on their desktops. Users have the option to run these models locally or deploy them on NVIDIA DGX Cloud or other accelerated cloud or data center infrastructure.
The DGX Spark and DGX Station bring the power of the Grace Blackwell architecture, previously only available in data centers, to desktop computers. Global system builders such as ASUS, Dell, HP Inc., and Lenovo have collaborated to develop the DGX Spark and DGX Station.
"AI has transformed every layer of the computing stack. It stands to reason a new class of computers would emerge—designed for AI-native developers and to run AI-native applications," said Jensen Huang, founder and CEO of NVIDIA. "With these new DGX personal AI computers, AI can span from cloud services to desktop and edge applications."
Igniting Innovation With DGX Spark
The DGX Spark is the world's smallest AI supercomputer, designed for researchers, data scientists, robotics developers, and students to explore generative and physical AI with high performance and capabilities.
At the core of DGX Spark is the NVIDIA GB10 Grace Blackwell Superchip, optimized for desktop use. The GB10 features a powerful NVIDIA Blackwell GPU with fifth-generation Tensor Cores and FP4 support, delivering up to 1,000 trillion operations per second of AI compute for fine-tuning and inference with the latest AI reasoning models.
The GB10 Superchip utilizes NVIDIA NVLink-C2C interconnect technology to provide a CPU+GPU-coherent memory model with 5x the bandwidth of fifth-generation PCIe, optimizing performance for memory-intensive AI developer workloads.
NVIDIA's full-stack AI platform allows DGX Spark users to seamlessly move their models between desktops, DGX Cloud, or other infrastructure with minimal code changes, simplifying the prototyping and iteration process.
Full Speed Ahead With DGX Station
The NVIDIA DGX Station brings data-center-level performance to desktops for AI development. It is built with the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip, offering a massive 784 GB of coherent memory space for large-scale training and inferencing workloads.
The GB300 Desktop Superchip features an NVIDIA Blackwell Ultra GPU with latest-generation Tensor Cores and FP4 precision, connected to a high-performance NVIDIA Grace CPU via NVLink-C2C for optimal system communication and performance.
Additionally, the DGX Station features the NVIDIA ConnectX-8 SuperNIC for hyperscale AI computing workloads, providing networking at up to 800 Gb/s for fast and efficient network connectivity.
By combining these capabilities with the NVIDIA CUDA-X AI platform, teams can achieve exceptional desktop AI development performance. Users also have access to NVIDIA NIM microservices through the NVIDIA AI Enterprise software platform, offering optimized inference microservices with enterprise support.
Availability
Reservations for DGX Spark systems are now open on nvidia.com. DGX Station is expected to be available from manufacturing partners like ASUS, BOXX, Dell, HP, Lambda, and Supermicro later this year.
