NVIDIA DGX Cloud Now Available for Broad Use
The NVIDIA DGX Cloud, a service that enables companies to become AI companies, is now widely accessible. Thousands of NVIDIA GPUs are available on Oracle Cloud Infrastructure, as well as NVIDIA infrastructure in the U.S. and U.K. DGX Cloud, introduced at NVIDIA's GTC conference in March, is an AI supercomputing service that provides enterprises with immediate access to the necessary infrastructure and software for training advanced models in generative AI and other groundbreaking applications.
"Generative AI has become a business necessity for leading companies in every industry, prompting many enterprises to seek faster computing infrastructure," said Pat Moorhead, chief analyst at Moor Insights & Strategy. According to recent estimates by global management consultancy McKinsey, generative AI could contribute over $4 trillion to the economy annually by transforming proprietary business knowledge across various industries into next-generation AI applications. Early pioneers in different industries are already driving transformative change using generative AI.
Healthcare companies are utilizing DGX Cloud to train protein models for accelerating drug discovery and clinical reporting with natural language processing. Financial service providers are using DGX Cloud to predict trends, optimize portfolios, build recommender systems, and develop intelligent generative AI chatbots. Insurance companies are developing models to automate claims processing. Software companies are leveraging DGX Cloud to create AI-powered features and applications. Additionally, DGX Cloud is being used to build AI factories and digital twins of valuable assets.
DGX Cloud instances offer dedicated infrastructure that enterprises can rent on a monthly basis, allowing customers to quickly and easily develop large, multi-node training workloads without waiting for high-demand accelerated computing resources. "The availability of NVIDIA DGX Cloud provides a new pool of AI supercomputing resources with nearly instant access," Moorhead stated. This simplified approach to AI supercomputing eliminates the complexity of acquiring, deploying, and managing on-premises infrastructure. DGX Cloud, equipped with NVIDIA DGX AI supercomputing and NVIDIA AI Enterprise software, enables businesses worldwide to access their own AI supercomputer through a web browser.
Each DGX Cloud instance is equipped with eight NVIDIA 80 GB Tensor Core GPUs, providing 640 GB of GPU memory per node. A high-performance, low-latency fabric ensures that workloads can scale across clusters of interconnected systems, allowing multiple instances to function as a massive GPU. DGX Cloud incorporates high-performance storage to offer a comprehensive solution.
Enterprises can manage and monitor DGX Cloud training workloads using NVIDIA Base Command Platform software. This platform offers a seamless user experience across DGX Cloud and on-premises NVIDIA DGX supercomputers, enabling enterprises to combine resources as needed. DGX Cloud also includes NVIDIA AI Enterprise, the software layer of the NVIDIA AI platform, which offers over 100 end-to-end AI frameworks and pretrained models to accelerate data science pipelines and streamline the development and deployment of production AI.
