Enterprises around the globe have been adopting hybrid for the past few years. While the reasons organizations come to hybrid are nothing new – from regulatory and data sovereignty requirements, limited connectivity or operating in an air-gapped environment -- hybrid has become increasingly central to businesses around the world.
We’re grateful for our OEMs, partners and customers who have been with us on this journey since 2017 when we launched Azure Stack Hub, and who continue to invest and innovate with us. We’re humbled to continue to experience first-hand the great work our customers and partners do across industries with the Azure Stack portfolio.
Our customers’ journeys speak for themselves, and today at Ignite, we’re sharing two examples of success that will resonate with companies exploring a hybrid approach. The first is Volkswagen Financial Services, which uses Azure Stack Hub to modernize their applications, and Oman Data Park, which empowers customers across Oman with their managed Azure Stack Hub offering.
From a product perspective, we’re delighted to deliver the general availability of GPU capabilities on Azure Stack Hub in collaboration with NVIDIA and AMD.
GPU-enabled NC and NV Series virtual machines (VMs) unlock the ability to run AI, machine learning (ML), VDI and Inferencing capabilities on-premises and at the edge. We offer three GPU models and associated VM sizes through NVIDIA V100 Tensor Core, NVIDIA T4 Tensor Core and AMD Mi25 GPUs. These physical GPUs align with the following Azure N-Series VM types, as follows:
- NCv3 (NVIDIA V100 Tensor Core GPU): These enable learning, inference, and visualization scenarios.
- NCasT4_v3 (NVIDIA T4 Tensor Core GPU): This new VM size enables light learning, inference, and visualization scenarios.
- NVv4 (AMD Mi25): These enable partitioned GPU VM sizes useful for visualization scenarios.
For Machine Learning workloads, we are bringing NVIDIA NGC to the Azure Stack Hub marketplace, with pre-deployed ML packages to help you get started in minutes. Now, you can train in the cloud with Azure Machine Learning and deploy on the Edge for inference.
Customers may purchase a solution with any of these GPUs through our supported hardware vendors.
“Lenovo and Microsoft’s partnership on Azure Stack Hub continues to grow by helping customers modernize their IT infrastructure. We were the first OEM partner to ship and deploy an Azure Stack Hub and are currently testing the next level of GPU capabilities that will further fuel innovation for high-end remote visualization and deep learning.“
Shekhar Mishra, Director, SDI solutions, Lenovo DCG
At Ignite this year, we're also bringing on Teradici as a VDI partner to make GPU-based VDI solutions production-ready.
“Teradici has collaborated closely with Microsoft, NVIDIA and AMD to validate Cloud Access Software on Azure Stack Hub, now with available GPU support. This ensures the newly launched solution provides seamless performance for its customers with a particular need for graphics-intensive applications, such as media production and computer-assisted design. The expanded capabilities of the new offering provide a valuable option for firms who prefer an on-premises solution to support high-performance workloads for secure, remote collaboration.” Read more here.
John McVay, Director, Strategic Alliances, Teradici
“Halliburton is collaborating with Microsoft to onboard their DecisionSpace® 365 Geosciences Suite on Azure Stack Hub, making use of the new GPU support available on the platform. This will enable Halliburton’s customers to run DecisionSpace® 365 cloud applications on iEnergy® Hybrid Cloud - from Azure Global to countries and regions where they need to run on-premises.”
Shashank Panchangam, Chief Product Manager, Cloud Services, Halliburton
Addenbrooke's Hospital recently started a collaboration with Microsoft Azure Stack Hub to demonstrate an end-to-end solution for an inferencing pipeline that accelerates a key part of cancer treatment for patients, which makes use of NVIDIA GPU V100 and T4. The robust inferencing pipeline code, based on Project InnerEye components, will enable them to run containerized models on VMs with 2 NVIDIA V100 Tensor Core GPUs as well as NVIDIA T4 Tensor Core, automating the selection and inference of the correct AI model for each CT scan.
“The collaboration between Project InnerEye and Addenbrooke’s Hospital exemplifies what can be achieved when NVIDIA’s technology is applied to solve real-world challenges in medical imaging.”
Mona Flores, Global Head of Medical AI at NVIDIA
Other News
We’re also excited to share with you the announcements down below and please make sure to check out the platform blog to learn more about our platform improvements on Azure Stack Hub. And, if you’re a developer, you may enjoy our developer blog post.
Deploy Azure Arc for Data Services - Public Preview on Azure Stack Hub
Operators can use Azure Stack and Azure Kubernetes Service (AKS)-engine Kubernetes clusters as the ideal infrastructure for hosting Azure Arc data services. Users can provision SQL Server and PostgreSQL server instances and use them in combination with Azure-consistent VMs, virtual networks, storage, functions, web apps and other resource types. Learn more https://aka.ms/ash-arc-for-data-services
Use Azure Kubernetes Services on Azure Stack Hub
We have launched the preview of the AKS Resource Provider on Azure Stack Hub. Built on top of the existing AKS engine, the resource provider will provide customers the ability to manage Kubernetes clusters from the portal and CLI in the same way as in Azure AKS. Join the preview here: https://aka.ms/ash-aks-private-preview
Join the preview of Azure Container Registry on Azure Stack Hub
At Ignite today, we’re also launching the preview of Azure Container Registry on Azure Stack Hub, providing customers the ability to store and retrieve container images from local container registries. Join the preview here: https://aka.ms/ash-acr-private-preview.
Next Steps
In the months and years to come, we look forward to partnering with you to transform your business. Learning from you, about how you use our technology to build solutions to achieve more, is what energizes us.
If you haven’t seen it, please also make sure to check out the Ignite session from Natalia Mackevicius about what’s new with Azure Stack.
Because Ignite is virtual this year, we created a mobile app that allows you to look at Azure Stack hardware and new form factors through augmented reality in the comfort of your environment. You can download it for your iOS or Android device or learn more about it here. We hope you enjoy it.