Nvidia DGX Cloud
Nvidia's internal cloud for AI development and operations
Nvidia DGX Cloud is the company's internal environment for building and operating artificial intelligence at scale. It supports demanding internal use cases, develops open source models, and validates new system architectures before external release.
- Supports training across tens of thousands of GPUs
- Serves as Nvidia's internal AI proving ground
- Runs across major cloud service providers
- Develops open source models like Nemotron
- Uses DSX OS for operational consistency
What it is
Nvidia DGX Cloud serves as the internal cloud environment for Nvidia. It is designed to build and operate artificial intelligence systems at scale. The platform supports the company's most demanding internal workloads, including the development of open source frontier models and foundational models. It also validates new system architectures and runs production AI tasks. The infrastructure relies on Nvidia accelerated computing resources. Operational patterns proven within this environment are applied across the wider ecosystem through partners.
The compute
The system runs on Nvidia accelerated infrastructure. It supports multi-node training across tens of thousands of graphics processing units in production. Nvidia DSX OS acts as the operating layer for these AI factories. This software helps partners bring infrastructure online and maintain runtime consistency. It automates fleet health checks and ensures reliable operation at scale. The Exemplar Cloud component improves performance per total cost of ownership through validated hardware and software stacks. Performance benchmarking tools ensure consistent GPU quality.
Where the GPUs are
DGX Cloud operates across cloud service providers and Nvidia Cloud Partners. This distribution ensures that software and operational patterns are applicable throughout the ecosystem. Specific partnerships include offerings with Amazon Web Services, Google Cloud, Microsoft Azure, and Oracle Cloud Infrastructure. These collaborations provide co-engineered accelerated computing clusters. The setup allows for flexible term lengths and access to Nvidia experts via each partner's marketplace or private offer channels.
Its place in AI
The platform acts as a proving ground for Nvidia. Operational challenges encountered at scale are resolved inside DGX Cloud first. Solutions are then transformed into repeatable software, architectures, and reference implementations. These components are externalized as modular, open infrastructure tools. Models such as Nvidia Nemotron are trained and optimized on this cloud. The Nemotron family demonstrates the scale and reliability of the AI factory. Other projects like Alpamayo for autonomous vehicles and Isaac GR00T for humanoid robots also leverage this infrastructure.
Questions people ask about Nvidia DGX Cloud
What is the primary purpose of Nvidia DGX Cloud?
It is Nvidia's internal cloud environment used to build and operate AI at scale. It supports demanding internal use cases, develops open source models, and validates new system architectures before they are released externally.
Which cloud providers host Nvidia DGX Cloud services?
The platform runs across various cloud service providers and Nvidia Cloud Partners. Specific integrations exist with Amazon Web Services, Google Cloud, Microsoft Azure, and Oracle Cloud Infrastructure, offering co-engineered clusters.
What software layer manages the AI factories in DGX Cloud?
Nvidia DSX OS serves as the operating layer. It helps partners bring infrastructure online, maintains runtime consistency, automates fleet health, and ensures reliable operation of AI infrastructure at scale.
Good reading and sources
There is no English Wikipedia article on Nvidia DGX Cloud yet, so the facts on this page come from the sources listed above: nvidia.com/en-us/data-center/dgx-cloud/. Some are the subject's own pages, so its claims about itself are reported as its claims. EveryNeocloud wrote the text in its own words, with AI assistance, and compares it with the sources to make sure no passage is copied.
It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.