At the March 21, 2023 GTC-NVIDIA announced a new development, the NVIDIA DGX™ Cloud. This is a supercomputing AI service that gives enterprises instant access to the infrastructure and software they need to train advanced generative AI models and other disruptive applications.
Dedicated clusters of NVIDIA DGX AI supercomputers, DGX Cloud provides paired with NVIDIA AI software. Enterprises lease these clusters on a monthly basis, thereby providing the ability to quickly and easily scale the development of large, multi-node training workloads without having to wait for accelerated computing resources.
Thus, a new service from NVIDIA allows every enterprise to access their own AI supercomputer with a simple web browser. The various complexities of purchasing, deploying and managing local infrastructure will now be eliminated.
NVIDIA is already working with leading cloud service providers to host DGX Cloud infrastructure, starting with Oracle Cloud Infrastructure (OCI). Its OCI supercluster provides a purpose-built RDMA network, bare-metal computing and high-performance on-premises and block storage that can scale to superclusters with more than 32,000 GPUs. Microsoft Azure is expected to begin providing DGX Cloud next quarter, and the service will soon expand to Google Cloud and other services.