Broadcom bolsters VMware Edge Compute Stack

A slew of updates in VMware Edge Compute Stack (ECS) 3.5 are aimed at helping customers more easily manage edge devices, applications, and infrastructure across multiple locations, and industry watchers say the updated platform will support AI at the edge.

Enterprises are expanding edge computing deployments to meet the scalability and performance requirements of AI applications, and tech vendors are extending feature sets to enable AI in edge locations, according to IDC, which predicts that edge computing investments will reach $232 billion in 2024. “Edge computing will play a pivotal role in the deployment of AI applications,” says Dave McCarthy, research vice president, cloud and edge services at IDC.

Broadcom, which acquired VMware in November 2023, has been adding to the edge management capabilities in the VMware ECS and VMware Edge Cloud Orchestrator (VECO) solutions. Updates in its latest offering will help customers simplify operations by reducing management complexity, more easily managing dispersed locations with limited IT resources, automating software patching, and gaining visibility into application and traffic behaviors at the edge with integrated telemetry.

“Edge computing continues to be an important aspect of infrastructure design to offset limitations in public cloud deployments. These limitations include network latency, the cost of data movement, regulatory compliance requirements, and overall business continuity. However, edge computing brings with it new challenges in terms of managing large amounts of remote infrastructure,” IDC’s McCarthy says. “VMware Edge Cloud Orchestrator aims to automate the management of large-scale edge deployments.”

To further simplify the deployment of applications and infrastructure across thousands of locations, ECS 3.5 includes new or updated features such as:

• Zero-touch orchestration, which simplifies deployment and application lifecycle management across multiple sites by leveraging GitOps and desired state management.
• Pull-based architecture, which puts less burden on the management plane, allowing it to achieve a much higher scale.
• Edge fleet management, which can manage large-scale edge infrastructure deployments of single hosts by providing streamlined visibility, deployment, operational processes, and automatic updates with the SaaS-based ECS service.
• SaaS delivery model, which enables Broadcom to introduce new VECO capabilities continuously.
• Decoupled management and control plane, which brings added deployment flexibility to ensure that the control plane to manage edge workloads is local to the edge location.
• Single-node updates, which support virtual machine and container workloads and also include single-node Kubernetes, making it feasible to scale to thousands of locations.
• Automatic updates, which alleviate the pain points associated with manual updates by ensuring the latest patches, compliance standards, and performance enhancements are applied.
• Edge infrastructure and application monitoring, which configures metrics gathering and transmission for infrastructure, virtual machines, and Kubernetes workloads—enabling customers to start monitoring edge components quickly.

“VECO features a pull-based architecture, which is ideal for customers with many edge sites and varying levels of connectivity. Similar to how an iPhone updates itself, edge sites update their applications and infrastructure when the edge host is connected and new updates or configuration changes are published,” according to a VMware blog announcing the product updates.

According to IDC’s McCarthy, upgrades to VMware edge products will help customers more easily manage large-scale, heterogeneous edge deployments. And VMware is looking to help customers manage whatever components they have in their edge infrastructure.

“As enterprises add edge infrastructure to existing cloud environments, they are looking for consistency. VMware is taking its successful software-defined data center concepts to create a software-defined edge. This benefits customers by leveraging similar tools and skills to manage infrastructure and applications wherever they are deployed,” McCarthy says. “Many edge offerings are tied to specific vendor ecosystems. For example, all of the major public cloud providers have edge management software, but it is limited to resources delivered by that cloud. VMware’s approach of supporting a variety of clouds and on-premises infrastructure means that it is applicable to more scenarios than the competition.”

And as AI continues to dominate technology conversations, the need to support AI at the edge increases. Enterprises are moving to a distributed data center model for real-time processing at the edge, and VMware should continue to invest in AI at the edge, according to McCarthy.

“As the AI conversation shifts from large training in the cloud to inference at the edge, VMware needs to stay on top of emerging edge AI needs from customers while continuing to build consistency between data center, cloud, and edge environments,” McCarthy says.

VMware enlisted more than 100 customers to participate in the technical preview of ECS 3.5, which is generally available now.