Platform and application virtualization are rapidly coming to the edge of the cloud. Paving the way are the multicore processors at the heart of devices for the most demanding “internet of things,” mobile and embedded applications.

As virtualization takes hold at the cloud’s edge, it’s finding its first applications in the industrial IoT. Virtualization allows a smart edge device, such as an industrial robot, to run virtual machines and containers to manage concurrently such artificial intelligence-driven application workloads as sensing, data ingestion, signal processing, computer vision, adaptive control and human-machine engagement.

In the process, edge-embedded hypervisors and other virtualization technologies can save space, use power more efficiently and streamline monitoring, diagnostics, troubleshooting and support requirements associated with managing intelligent systems.

More broadly, virtualization technology can ensure that two or more edge devices can be elastically and adaptively combined to handle dynamic workloads. It can enable those devices to adapt seamlessly and in real time to the peaks and lulls of many edge application scenarios, such as the industrial IoT and real-time congestion management in private, public and hybrid clouds. And it can boost cloud-to-edge infrastructure efficiency by helping to allocate workloads dynamically across intermediate layers of computation, networking and storage.

As VMworld 2018 nears, Wikibon expects VMware Inc. to put greater emphasis on opportunities in cloud-to-edge virtualization. Key to this push is NSX Data Center, which is the centerpiece of VMware’s software-defined networking strategy. It is VMware’s top priority for multicloud and edge deployments. In the past two years, the company has been steadily expanding NSX’s reach, scale and simplicity all the way to the edge.

NSX now integrates with VMware’s vSphere server virtualization, ESXi bare-metal hypervisor and growing portfolio of containerized microservices offerings. Going forward, the company’s strategic focus is on making NSX platform services consumable at the software layer, with more routing, load-balancer, firewall and traffic microsegmentation functions moving into that layer.

This intelligent networking is necessary to expand VMware’s software-defined data center reach so that it can move computation transparently and securely to and among IoT edge devices. It will enable VMware users to treat a collection of IoT devices as a virtualized data center running diverse VMs, containerized microservices and other workloads within a cloud-to-edge computing fabric.

Over the past two years, VMware has made a series of significant moves that hint at a more comprehensive software-defined edge strategy to come. Key recent milestones in this regard include releases of the following solutions:

• NSX for vSphere 6.3: This upgrade to VMware’s core server virtualization platform enables users to run application workloads in virtualized data centers that are distributed across hybrid clouds.
• NSX-T1.1: This new product enables virtual networking in heterogeneous cloud-native environments, including OpenStack’s Newton, and also manages KVM hypervisors from Canonical and Red Hat. In a subsequent enhancement following its initial release, NSX-T added support for managing heterogeneous hypervisors, bare metal and container platforms, including sister company Pivotal’s Cloud Foundry platform-as-a-service containerized platform.
• NSX Hybrid Connect: This new product enables customers to migrate workloads from any VMware environment to a modern software-defined data center environment running on-premises, in the public cloud or operated by a VMware cloud partner, with a single point of control via the vRealize cloud management and optimization suite.
• NSX SD-WAN: This is a rebranding of software-defined wide area network technology that VMware acquired with VeloCloud Networks last year. The acquisition added to VMware’s network virtualization portfolio in support of software-defined connection of enterprise data centers to branch offices over large geographic distances. It provides a connectivity layer to connect IoT edge devices, branch offices, hybrid clouds and networks, and users into a consistent framework. It has the ability to perform traffic microsegmentation, so that, for example, Wi-Fi bandwidth consumed by IoT edge devices in a branch can be managed independently of corporate traffic. It abstracts the networking away from its underlying hardware and software environments, and it’s being integrated with NSX Data Center and VMware NSX Cloud virtual network to enable enforcement of consistent networking and security policies from the data center to the branch to the cloud with central visibility and control. The VeloCloud acquisition was the latest in VMware’s buildout of its software-defined networking portfolio, following its startup Nicira Inc. and PLUMgrid Inc. acquisitions in 2012 and 2016, respectively.
• vCloud NFV-OpenStack Edition 3.0: The new vCloud NFV-OpenStack Edition 3.0 infrastructure platform facilitates deployment of virtualized network functions to the edge across 5G and distributed computing environments within multicloud telco networks. It integrates with VMware’s NSX-T Data Center offering in support of network virtualization and microsegmentation platform for hypervisor environments, container deployments and native workloads running in public cloud environments. It also includes a data-plane acceleration feature that helps reduce processing requirements in network edge environments.

Just as important, VMware over the past year has expanded the edge virtualization focus of its partnerships. The key partnership announcements in this push are the following:

• SAP and VMware for strong device deployment, security and management: Last October, VMware and SAP announced that they’re partnering to make enterprise IoT deployments easier. They have teamed to launch a new enterprise-focused IoT initiative that encompasses devices, the cloud, infrastructure and other applications. They are focusing on fast deployment, scalability and on-schedule execution of enterprise IoT projects. They plan to offer customers new IoT services that incorporate SAP’s analytics, business transaction and device data with VMware’s operational systems. Their joint solution will enable rapid deployment of IoT architecture across the cloud, gateways and micro data centers with strong security and management capabilities.
• Microsoft, Dell and VMware for centralized cloud-to-edge device and application provisioning, security and management: This past May, Dell and VMware announced a partnership with Microsoft to offer a forthcoming IoT edge-device management solution streamlined cloud-to-edge device management, centralized monitoring and security. It will focus on industrial IoT applications such as predictive maintenance and supply chain visibility. Microsoft Azure IoT Edge will provide the intelligence to manage IoT devices need to manage device-level software individually and in coordination with each other. Though it’s optimized for Dell’s Atom Edge Gateway hardware, it will also be able to manage, monitor and secure other vendors’ gateways and edge systems. VMware’s Pulse IoT Center will provide secure, enterprise-grade IoT device onboarding, management, monitoring and security, and will interface the Dell and other gateways to the Microsoft cloud-based service.

Increasingly, Dell/VMware’s server, network and device virtualization platforms — including VSphere, NSX and Pulse — will manage diverse real-time application workloads. These platforms will need to play together seamlessly in virtual cloud-to-edge fabrics to ensure that AI and other workloads — running in VMs, containers and serverless environments — can run in parallel on the edge devices with assured service levels managed across the distributed edge.

Wikibon expects VMware to deepen its focus on Pulse IoT Center in its edge virtualization roadmap. The platform, which was introduced last year, provides enterprise information technology professionals with visibility into IoT infrastructure performance metrics, such as CPU utilization and battery life, and correlate it with historic data, as well as information from other devices, to help identify anomalies and optimize device performance. It incorporates Liota, which is a client-side module that resides on edge gateways or can be embedded in devices, enabling the transmission of edge operational data — such as that associated with performance of device-embedded AI workloads — for centralized management of cloud-to-edge optimization.

For VMware parent Dell Technologies Inc., partnerships are important in its distributed-core SDDC strategy, as Michael Dell, the company’s founder and chief executive officer, said during an interview on theCUBE, SiliconANGLE’s mobile TV studio at Dell Technologies World conference in May. Dell and its subsidiary VMware are positioning themselves as the premier infrastructure providers for compute power at the network edge, especially for AI in IoT devices.

“We’ve positioned ourselves as the central infrastructure company,” said Dell. “We’re on the [network] edge, [we] have a distributed core.”

You can catch Dell and other Dell EMC, VMware and partner executives interviewed live on theCUBE from VMworld 2018, which takes place Aug. 27-29 in Las Vegas.

Image: Duncan Hill/Flickr