NVIDIA DGX Spark Founders Edition — Personal AI Supercomputer for AI Development
Build. Experiment. Fine-tune. Run AI — right from your desk.
As AI models become increasingly capable and complex, developers need more compute and memory to experiment, fine-tune and run modern AI workloads.
The NVIDIA DGX Spark is designed to provide that capability locally, combining the NVIDIA Blackwell GPU architecture with a high-performance Grace CPU and 128GB of coherent unified system memory.
With up to 1 petaFLOP of FP4 AI performance, DGX Spark provides a powerful platform for developing and testing AI applications without relying entirely on remote cloud infrastructure.
Why DGX Spark
- Scale to two. Built-in NVIDIA ConnectX-7 networking links two DGX Spark systems for inference on models of up to 405 billion parameters².
- Ready out of the box. Ships with NVIDIA DGX OS (Ubuntu Linux) and the NVIDIA AI software stack preinstalled. It works with familiar tools such as PyTorch, Jupyter and Ollama, and gives access to NVIDIA NIM and NVIDIA Blueprints through the NVIDIA Developer Program.
- Develop locally, deploy anywhere. Because it runs on the same NVIDIA AI platform architecture, work moves from your desk to DGX Cloud or any accelerated data centre or cloud with minimal rework.
- Built for agent development. NVIDIA OpenShell, an open-source runtime, supports building always-on autonomous agents more safely on DGX Spark.
Get yours while stocks last
Quantities are limited. To reserve a unit, request a quotation or discuss deployment for your team, contact NovaGlobal:
📧 [email protected] · 📞 +65 6465 0204
