Artificial Intelligence (AI) is being used more and more now a days because it helps organizations gain insights which they can apply to accelerate innovations and business results. Having an AI-first infrastructure is critical to develop and deploy AI applications.
Microsoft Azure along with NVIDIA will deliver purpose-built AI supercomputers in the cloud to handle the most demanding real-world workloads while also meeting price/performance and time-to-solution requirements. They come with advanced machine learning tools that will help you to incorporate AI into your own work, making your simulations smarter and your decisions more intelligent.
This computer will be powered by Microsoft Azure’s advanced supercomputing infrastructure and NVIDIA GPUs. It will help enterprises train, deploy, and scale AI, including large, state-of-the-art models. Azure’s cloud-based AI supercomputers comprise powerful and scalable ND- and NC-series virtual machines that are optimized for AI distributed training and inference. NVIDIA will make use of Azure’s scalable virtual machine instances to research and accelerate advances in generative AI.
Microsoft and NVIDIA will also work together to make Microsoft’s DeepSpeed deep learning optimization software better. NVIDIA will make its full stack of AI workflows and software development kits available to Azure enterprise customers.
Scalable peak performance for AI training with NVIDIA Compute and Quantum-2 InfiniBand on Azure
Microsoft Azure’s virtual machine instances that are optimized for AI use NVIDIA’s most advanced data center GPUs. These instances are the first public cloud instances to have NVIDIA Quantum-2 400Gb/s InfiniBand networking. It allows customers to deploy thousands of GPUs in a single cluster for training large language models, building complex recommender systems and enabling generative AI.
The current Azure instances have NVIDIA Quantum 200Gb/s InfiniBand networking and NVIDIA A100 GPUs. The future instances will have NVIDIA Quantum-2 400Gb/s InfiniBand networking and NVIDIA H100 GPUs. These new instances will be integrated with Azure’s advanced compute cloud infrastructure, networking, and storage to offer scalable peak performance for AI training and deep learning inference workloads of any size.
Microsoft Azure’s AI-first cloud infrastructure and toolchain featuring NVIDIA are making a big impact in retail. With a GPU-accelerated computing platform, customers can quickly churn through models to determine the best-performing model.
Accelerated AI development and deployment
The cloud-based AI supercomputer will support a wide range of AI applications and services, including Microsoft DeepSpeed and the NVIDIA AI Enterprise software suite.
Microsoft DeepSpeed will use the NVIDIA H100 Transformer Engine to speed up transformer-based models used for large language models, generative AI, and writing computer code, along with other applications. This technology applies 8-bit floating point precision capabilities to DeepSpeed to dramatically increase the speed of AI calculations for transformers.
NVIDIA AI Enterprise is a globally adopted software of the NVIDIA AI platform. It is certified and supported on Microsoft Azure instances with NVIDIA A100 GPUs. Support for Azure instances with NVIDIA H100 GPUs will be added in the future.
NVIDIA AI Enterprise includes the NVIDIA Riva for speech AI and NVIDIA Morpheus cybersecurity application frameworks. This helps streamline every step of the AI workflow, ranging from data processing and AI model training to simulation and large-scale deployment.