Advanced Computing in the Age of AI | Saturday, December 3, 2022

Lenovo Links with Azure HCI Stack for Edge-Core-Cloud Compute, Storage 

Lenovo has linked with Microsoft to offer edge computing and storage hardware for organizations deploying Azure cloud-based hyperconverged infrastructure (HCI) strategies. The devices, aimed at supporting IoT solutions and 5G technologies, is designed to address the increase in connected devices and the explosion of data created, analyzed, stored and managed from the edge to the core.

On the compute side, Lenovo’s ThinkAgile MX1021 server utilizes Microsoft Azure Stack HCI solutions to move compute to where data is created at the edge. Coming in a 1U height, half-width form factor, the server utilizes Intel Xeon D-2100 Series processors with between eight and 16 cores. It has 4 DIMM slots for up to 256GB of RAM, and 16TB of internal solid-state storage. Available as a 2-node Azure Stack HCI cluster, the ThinkAgile MX1021 handles temperatures from 0° to 55°C, according to Lenovo, along with high-dust and vibration.

On the storage side, the ThinkSystem DM7100 is an all-flash array and hybrid flash enterprise-class device with end-to-end NVMe technology and integrated hybrid cloud capability for Microsoft Azure and other public cloud platforms, according to Lenovo. The company said all-flash with integrated Azure cloud storage tiering “enables customers to deploy applications that leverage Azure Cloud Services with consistent data management across the enterprise.” The device’s data reduction capabilities will be retained in the cloud, Lenovo said, reducing cloud footprints by up to 66 percent, while NVMe can improve data analytics speed by 30 percent.

Kamran Amini, VP/GM of Server, Storage and Software Defined Infrastructure at Lenovo Data Center Group, told us Lenovo is the first server vendor to “extend the Microsoft HCI stack to the edge.” He also said Microsoft CEO Satya Nadella jumped the gun on today’s announcement, discussing it publicly at last November’s Microsoft Ignite, the company’s annual conference.

“I think Satya hinted at it, and other executives posted Satya’s main keynote,” Amini said, “he emphasized that if you want to run Microsoft HCI and you want to run at the edge, the only platform you really want to run that on is Lenovo's.”

The advantage of the Lenovo-Azure offering is “you have a single point of contact, single point of support, single validated design. And … now they can extend that to the edge … and seamlessly have that all be integrated if they want to port it to the public cloud with Azure. So this is where we're uniquely positioned today.”

He cited the example of retailers that typically have core enterprise data centers and a small-scale “data center” at their retail stores for localized compute. “How can I extend that in a hyperconverged way? I have deployed the core, now I want to run that same hyperconverged capability on a different type of server that is meant for the edge, with a security capability, connectivity capability, with an accelerator to do global AI on the edge?”

Similar deployments can be made at local factories that are part of larger manufacturing companies and schools within larger educational institutions. Key to the strategy, Amini said, “is if it's a Microsoft environment … So that's where I think the edge aspects going to play. I think Microsoft, purely because of the ease of tapping into their public cloud, provides customers great choice, they can run everything in a hybrid form at every level, in my core, even at my edge if I need to, and in a hybrid landscape.”

While noting that edge computing is not new, Charles King of Pund-IT Reviews said that “for decades, enterprises of all kinds have deployed and often struggled to support computing processes at the far edges of their networks in locations, including retail outlets, manufacturing facilities, healthcare clinics and remote and branch offices.”

The difficulty, he said, is that “solutions for edge of network computing varied widely in quality and performance, required complex integration to be compatible with primary data centers and were difficult to manage and maintain. In addition, moving substantial quantities of data across networks for further analysis in primary data centers was costly and time consuming.”

“Thankfully, the appearance of new solutions, including … HCI and integrated public cloud solutions have improved the situation considerably,” King said. “Lenovo’s new and updated edge-to-cloud ThinkAgile HCI and ThinkSystem storage solutions, which optimally leverage Microsoft Azure’s cloud services, are good examples of how vendors are helping their customers gain the full advantage of enterprise edge computing.”

Add a Comment