Dell Technologies has aggressively promoted Azure Stack, Microsoft’s software package that allows enterprises to run a complete copy of the Azure cloud service within their own data center.
Now it has introduced a hyperconverged infrastructure (HCI) system designed to support Azure Stack: a 1U server that allows organizations to start small with their deployment and grow.
Formally known as Dell Integrated System for Microsoft Azure Stack HCI, the single-node system is designed for customers with smaller data-center footprints, but is expandable to support AI/ML workloads.
Up to now, Azure Stack HCI nodes were sold at least in pairs, and Dell priced the hardware based on the number of nodes and the components within it. “So if you’re buying half the nodes, that would be half the price,” said Shannon Champion, vice president of product marketing for HCI products at Dell.
Dell has integrated its Dell OpenManage software with Windows Admin Center and Azure Arc, Azure’s node-management and security system. Arc manages HCI nodes and enforces regulatory compliance, and Champion said many Azure Stack customers use it to keep resources on-prem.
“The idea of Azure Stack HCI is that these are workloads that cannot run in public cloud,” she said. “And so you have a choice when you decide what workload can go to Azure. If it can’t go into the public cloud then it would stay on premises and go to Azure Stack HCI OS.”
The new HCI system also comes with expanded GPU support. Customers have the option of getting Nvidia A30 or A2 GPU cards with their servers. Both are based on the Ampere architecture. Champion said that the A2 is suited for data visualization and for AI inferencing, while the A30 is a bigger, more powerful card more appropriate for high-end computing and AI training.
Dell Integrated System for Microsoft Azure Stack HCI is available now either for outright purchase or via Dell’s Apex consumption service.