Darius Baruo
Oct 13, 2025 18:19
NVIDIA CEO Jensen Huang delivers a DGX Spark AI supercomputer to Elon Musk at SpaceX’s Starbase, marking a brand new period in AI know-how with petaflop efficiency.
The panorama of synthetic intelligence is poised for a big leap as NVIDIA founder and CEO Jensen Huang personally delivered the cutting-edge DGX Spark AI supercomputer to Elon Musk at SpaceX’s Starbase, Texas. This supply marks the primary deployment of the world’s smallest AI supercomputer, which boasts a petaflop of efficiency and suits comfortably on a desk, based on NVIDIA Newsroom.
AI Supercomputing Energy
The DGX Spark is provided with 128GB of unified reminiscence, enabling it to run fashions with as much as 200 billion parameters regionally. This technological marvel is designed for builders, researchers, and creators who require supercomputer-class efficiency in a conveyable format. The system’s introduction comes 9 years after NVIDIA’s preliminary foray into AI with the DGX-1, extending the corporate’s affect past conventional knowledge facilities.
Strategic Location and Timing
The handoff coincided with the eleventh check of SpaceX’s Starship, essentially the most highly effective launch car thus far, highlighting the strategic significance of AI in area exploration and know-how improvement. By inserting the DGX Spark on the coronary heart of SpaceX’s operations, NVIDIA is positioning itself on the forefront of AI-driven innovation in aerospace.
Widespread Availability
Past SpaceX, the DGX Spark is predicted to achieve numerous modern hubs worldwide, from robotics labs to inventive studios. Notable locations embrace Ollama in Palo Alto, Arizona State’s robotics lab, and the studio of artist Refik Anadol, amongst others. Every supply is anticipated to contribute considerably to the development and accessibility of AI applied sciences globally.
NVIDIA plans to make the DGX Spark usually accessible beginning October 15, providing it by means of NVIDIA.com and its companions worldwide. This transfer is ready to democratize entry to petaflop AI capabilities, bringing high-performance computing inside attain of a broader viewers.
Picture supply: Shutterstock
