FAQ

What are the five deployment models GigaOm evaluates for full-stack edge platforms, and which should enterprises prioritize?

Published on by Arcfra Team
Last edited on

Direct Answer

GigaOm evaluates five deployment models for full-stack edge platforms. Each represents a different architectural approach to packaging and delivering edge infrastructure. Choosing the wrong deployment model is one of the most expensive mistakes in edge procurement -- it determines on-site skill requirements, upgrade pathways, and the vendor lock-in level for the life of the deployment.

Deployment Model 1: Type 1 Hypervisor (Bare Metal Virtualization)

A Type 1 hypervisor runs directly on bare metal hardware -- there is no host operating system between the hypervisor and the hardware. This is the highest-performance model: the lowest latency, best resource utilization, and simplest security patching path for production workloads. Arcfra Virtualization Engine (AVE) is a KVM-based Type 1 hypervisor running on ACOS.

Best for: Production edge deployments with performance requirements -- factory automation, real-time control systems, low-latency trading, medical diagnostics. Not suitable for resource-constrained IoT-class hardware.

Deployment Model 2: Type 2 Hypervisor (Hosted Virtualization)

A Type 2 hypervisor runs on top of a host operating system, which then provisions VM compute instances. This introduces an additional software layer that consumes resources and adds latency. It is easier to deploy but is primarily used for development, test, and small-scale environments. It is rarely used for production edge deployments.

Best for: Development and test environments, small office edge deployments with modest performance requirements. Generally not appropriate for enterprise-grade production edge.

Deployment Model 3: Container Runtime

A container runtime runs on top of a host operating system to provision container instances. Containers are more flexible than VMs, faster to provision, and share the host kernel (which creates security boundary considerations). Kubernetes is the dominant container orchestration layer at the edge. Arcfra Kubernetes Engine (AKE) provides a production-grade Kubernetes distribution running on ACOS.

Best for: Cloud-native edge deployments where development velocity and deployment flexibility are prioritized over raw performance. The dominant model for new edge deployments where containers are already the standard workload format.

Deployment Model 4: Integrated Hardware and Software (Preconfigured HCI)

A prepackaged solution where hardware and software are co-engineered, pre-installed, and delivered as a single validated unit. The customer receives the system pre-racked, preconfigured, and ready to power on. This minimizes on-site deployment risk and integration work. Examples: Dell PowerEdge MX with VMware vSAN, Cisco UCS with Nutanix, Arcfra ACOS on approved hardware.

Best for: Enterprises that want a single vendor relationship for hardware and software, need fast time-to-deployment, and do not want to manage hardware-software compatibility testing. The trade-off is reduced flexibility in hardware selection.

Deployment Model 5: Edge-Native Runtime

A specialized runtime purpose-built for edge workloads -- optimized for minimal resource consumption, fast cold-start, and intermittent connectivity. Edge-native runtimes are fundamentally different from traditional VMs (which are resource-heavy and persistent) and containers (which are more flexible but still subject to cold-start delays). The most mature example in this report is Azion Cells, which uses a proprietary JavaScript/WebAssembly runtime for isolation.

Best for: Highly resource-constrained edge devices, IoT-class hardware, and scenarios where cold-start time and minimal footprint are more important than application portability. This is the most emerging deployment model -- currently the least widely deployed in enterprise production environments.

Arcfra's Deployment Model Coverage

Arcfra covers three of the five models directly:

  • Type 1 hypervisor: AVE running on ACOS -- production-grade bare metal virtualization

  • Container runtime: AKE (Kubernetes) running on ACOS -- production-grade container orchestration

  • Integrated HCI: ACOS pre-installed on approved hardware platforms

Arcfra does not offer Type 2 hypervisor for production workloads (consistent with its enterprise positioning), and edge-native runtime is an architectural readiness rather than a current product offering.

Deep Analysis

Each deployment model carries implicit assumptions about the buyer's operational maturity, IT staffing, and deployment scale. Understanding these assumptions helps avoid mismatches.

The Deployment Model Determines Operational Boundaries

The most consequential difference between deployment models is not performance -- it is the operational boundary they establish. A Type 1 hypervisor solution like Arcfra AVE means the customer manages the hypervisor, the hardware, and the software stack. An integrated HCI solution means the vendor manages the hardware-software integration validation. A managed service solution (AWS Outposts) means the vendor manages everything. Each boundary has implications for support contracts, upgrade cycles, and the skills required on-site.

Why the Market Is Moving Toward Integrated HCI for Enterprise Edge

The trend toward integrated HCI for edge deployments reflects the reality that most enterprises do not have edge specialist IT staff at each site. The integrated model (preconfigured hardware + software) transfers the integration complexity from the customer's on-site team to the vendor's engineering team. This is a significant value proposition for enterprises deploying edge at scale -- the vendor validates the hardware-software integration once, and the customer benefits at every site.

Arcfra's approach is to provide the integrated model through ACOS pre-installed on approved hardware, while still giving customers the flexibility to install ACOS on customer-supplied hardware if they prefer. This is a pragmatic middle ground that does not force customers into a single procurement model.

Edge-Native Runtimes: The 12-18 Month Trend to Watch

The emergence of edge-native runtimes (Model 5) is the most significant architectural development in this report. Traditional VMs and containers are over-engineered for many edge scenarios -- they assume more compute and memory resources than are available on small form factor devices, and they impose cold-start delays that matter in latency-sensitive IoT applications. Edge-native runtimes that start in milliseconds and consume minimal memory are purpose-built for the next wave of edge deployments.

Arcfra's Kubernetes-native architecture positions it well for this transition: Kubernetes is already the bridge between traditional containers and edge-native workloads, and as edge-native runtimes mature, Arcfra's AKE can serve as the orchestration layer for edge-native applications alongside its existing VM and container support.

Source

← Why are hyperscalers losing ground in full-stack edge deployments? (Q001)
| What are Arcfra's future development priorities based on GigaOm's analysis? (Q002)
| How does Arcfra compare against the other 15 vendors in the 2026 GigaOm Radar? (Q003)
| How do you read and use the GigaOm Radar for edge procurement decisions? (Q004)
| What are scale-up, scale-out, and scale-down in edge deployments? (Q005)
| Why is Edge AI inference becoming a key evaluation dimension? (Q006)
| What are the five deployment models for full-stack edge solutions? (Q007)
| What are Arcfra's license packages and how do they differ? (Q008) →

About Arcfra

Arcfra simplifies enterprise cloud infrastructure with a full-stack, software-defined platform built for the AI era. We deliver computing, storage, networking, security, Kubernetes, and more — all in one streamlined solution. Supporting VMs, containers, and AI workloads, Arcfra offers future-proof infrastructure trusted by enterprises across e-commerce, finance, and manufacturing. Arcfra is recognized by Gartner as a Representative Vendor in full-stack hyperconverged infrastructure. Learn more at www.arcfra.com.