NVIDIA DGX Spark AI Supercomputer
The power of NVIDIA Grace Blackwell on your desk.
Experience next-gen AI acceleration with NVIDIA Grace Blackwell architecture.
Spark Something Big
NVIDIA DGX Spark is a personal AI supercomputer designed to accelerate AI development, experimentation, and deployment. Built for businesses, developers, and AI researchers in the UAE, it brings next-generation AI computing to the desktop for generative AI, machine learning, and large language model development. Powered by the NVIDIA GB10 Grace Blackwell Superchip, DGX Spark delivers up to one petaFLOP of FP4 AI performance in a compact, power-efficient form factor. With 128 GB of memory, the NVIDIA AI software stack, and tools like NVIDIA NemoClaw, developers can prototype, fine-tune, and deploy advanced AI models and build safer autonomous AI agents.
Built for Next-Generation AI Workloads
Designed for AI developers, researchers, and enterprises, NVIDIA DGX Spark brings powerful AI computing to the desktop. With NVIDIA Grace Blackwell architecture, advanced AI software tools, and high-performance computing capabilities, DGX Spark helps accelerate generative AI development, AI model experimentation, and intelligent application building.
If you primarily work on another system, such as a laptop, and want to use your DGX Spark as a remote resource, this playbook shows you how to connect and work over SSH. With SSH, you can securely open a terminal session or tunnel ports to access web apps and APIs on your DGX Spark AI SuperComputer from your local machine.
There are two approaches: NVIDIA Sync (recommended) for streamlined device management, or manual SSH for direct command-line control.
NVIDIA DGX Spark
A Grace Blackwell AI Supercomputer on your desk.
NVIDIA GB10 Superchip: Experience up to 1 petaFLOP of AI performance at FP4 precision with the NVIDIA Grace Blackwell architecture.
128 GB of Coherent Unified System Memory: Run AI development and testing workloads with AI models up to 200 billion parameters at your desktop with a large, unified system memory.
NVIDIA ConnectX Networking: NVIDIA ConnectX high-speed interconnect enables dual DGX Spark scaling for AI models up to 405B parameters.
NVIDIA AI Software Stack: Full-stack generative AI—powered by NVIDIA tools, frameworks, libraries, and NVIDIA NIM.
Before you get started,
There are some important concepts to understand:
•
Secure Shell (SSH) is a cryptographic protocol for securely connecting to a remote computer over an untrusted network. It lets you open a terminal on your DGX Spark AI SuperComputer as if you were sitting at it, run commands, transfer files, and manage services—all encrypted end-to-end.
•
SSH tunneling (also called port forwarding) securely maps a port on your laptop (for example, localhost:8888) to a port on the DGX Spark where an app is listening (such as JupyterLab on port 8888). Your browser connects to localhost, and SSH forwards the traffic through the encrypted connection to the remote service without exposing that port on the wider network.
•
mDNS (Multicast DNS) lets devices discover each other by name on a local network without needing a central DNS server. Your DGX Spark AI SuperComputer advertises its hostname via mDNS, so you can connect using a name like spark-abcd.local (note the .local suffix), rather than looking up its IP address.
What you’ll accomplish
You will establish secure SSH access to your DGX Spark device using either NVIDIA Sync or a manual SSH configuration. NVIDIA AI SuperComputer Sync provides a graphical interface for device management with integrated app launching, while manual SSH gives you direct command-line control with port forwarding capabilities. Both approaches enable you to run terminal commands, access web applications, and manage your DGX Spark remotely from your laptop.
