Hypervisors Explained: Type 1 vs Type 2 Virtualisation
Virtualisation has transformed modern IT infrastructure, allowing organisations to run multiple operating systems on a single physical machine. At the heart of every virtualised environment is the hypervisor. This article explains the two types of hypervisors, compares the leading platforms, and helps you decide which is right for your needs.
What Is Virtualisation?
Virtualisation is the process of creating virtual (rather than physical) versions of computing resources — most commonly entire computers. A single physical server can host multiple virtual machines (VMs), each running its own operating system and applications in complete isolation. To the software inside each VM, it appears to be running on dedicated hardware, even though it is sharing the physical resources of the host.
The benefits are substantial: better hardware utilisation (typical physical servers run at 10-15% capacity; virtualised servers can reach 60-80%), rapid provisioning of new environments, simplified disaster recovery through VM snapshots and replication, and reduced physical footprint meaning less power, cooling, and rack space. Virtualisation is not new — the concept dates back to IBM mainframes in the 1960s — but it became mainstream in x86 server environments in the early 2000s and is now the default approach for almost all server deployments.
What Is a Hypervisor?
A hypervisor (also called a Virtual Machine Monitor or VMM) is the software layer that creates and manages virtual machines. It sits between the physical hardware and the virtual machines, allocating CPU, memory, storage, and network resources to each VM. The hypervisor is responsible for ensuring that VMs are isolated from each other — a crash or security breach in one VM should not affect others running on the same host.
Hypervisors are classified into two types based on where they sit in the software stack: Type 1 (bare-metal) and Type 2 (hosted). The distinction has significant implications for performance, security, and use cases.
Type 1: Bare-Metal Hypervisors
A Type 1 hypervisor runs directly on the physical hardware — there is no underlying operating system between the hypervisor and the hardware. The hypervisor is the operating system, purpose-built for one job: managing virtual machines. This direct hardware access gives Type 1 hypervisors superior performance, lower overhead, and better security compared to their Type 2 counterparts.
The leading Type 1 hypervisors include:
- VMware ESXi: The industry standard for enterprise virtualisation, part of the vSphere suite. ESXi has a tiny footprint (installs on a USB drive or SD card) and is managed through vCenter Server. VMware's ecosystem of tools for networking (NSX), storage (vSAN), and automation is unmatched, though licensing costs — particularly since the Broadcom acquisition — have driven many organisations to evaluate alternatives.
- Microsoft Hyper-V: Built into Windows Server and also available as a free standalone product (Hyper-V Server). Tightly integrated with the Windows ecosystem, Active Directory, and System Center. An excellent choice for Windows-centric environments, and its licensing model (per-core rather than per-VM) can be more cost-effective.
- Proxmox VE: An open-source virtualisation platform based on KVM and LXC containers. Proxmox has gained enormous popularity in the SMB and homelab space due to its zero licensing cost, capable web interface, and support for both VMs and containers. Enterprise support subscriptions are available for production use.
- KVM (Kernel-based Virtual Machine): Built into the Linux kernel itself, KVM turns any Linux installation into a hypervisor. It is the foundation for many cloud platforms (including Google Cloud and AWS's Nitro system) and enterprise products like Red Hat OpenShift Virtualisation. KVM offers exceptional performance and flexibility but requires Linux administration skills.
Type 2: Hosted Hypervisors
A Type 2 hypervisor runs as an application on top of a conventional operating system (Windows, macOS, or Linux). The host OS manages the hardware, and the hypervisor relies on it for device drivers, memory management, and scheduling. VMs are created within the hypervisor application, and each one runs as a process within the host OS.
The most common Type 2 hypervisors are:
- Oracle VirtualBox: Free and open-source, available on Windows, macOS, and Linux. VirtualBox is the go-to choice for developers and students who need to run different operating systems for testing, learning, or cross-platform development. It supports snapshots, shared folders, and a wide range of guest operating systems.
- VMware Workstation (Windows/Linux) and Fusion (macOS): Commercial products offering more advanced features than VirtualBox, including better performance, USB 3.0 passthrough, encrypted VMs, and integration with VMware's enterprise ecosystem. Popular with developers who need to test in environments that closely mirror production vSphere setups.
- Parallels Desktop: A macOS-specific product optimised for running Windows on Apple hardware, including Apple Silicon (M-series) Macs. It is the most seamless Windows-on-Mac experience, with Coherence mode that integrates Windows applications directly into the macOS desktop.
Because Type 2 hypervisors run on top of a full operating system, they inherit all of that OS's overhead — the host OS consumes CPU, memory, and disk I/O that could otherwise be allocated to VMs. There is also an additional layer of abstraction between VMs and hardware, which adds latency. For these reasons, Type 2 hypervisors are not suitable for production server workloads.
Type 1 vs Type 2 Comparison
Type 1 vs Type 2 Hypervisor Comparison
| Feature | Type 1 (Bare-Metal) | Type 2 (Hosted) |
|---|---|---|
| Runs On | Directly on hardware | On top of a host OS |
| Performance | Near-native (2-5% overhead) | Moderate (10-20% overhead) |
| Security | Smaller attack surface | Inherits host OS vulnerabilities |
| Use Case | Production servers, data centres | Development, testing, learning |
| Examples | ESXi, Hyper-V, Proxmox, KVM | VirtualBox, VMware Workstation |
| Management | Web/dedicated console | Desktop application |
| Scalability | Hundreds of VMs per host | A handful of VMs typically |
| Cost | Free (Proxmox/KVM) to expensive (VMware) | Free (VirtualBox) to moderate |
| Hardware Support | Limited to certified/tested hardware | Leverages host OS driver ecosystem |
Performance and Overhead
Modern Type 1 hypervisors leverage hardware-assisted virtualisation (Intel VT-x and AMD-V) to achieve near-native performance. CPU overhead is typically 2-5%, and with technologies like SR-IOV for network I/O and direct device passthrough, even I/O-intensive workloads perform well in VMs. Type 2 hypervisors also use hardware-assisted virtualisation but suffer from the additional layer of the host OS competing for resources. Disk I/O in particular can be significantly slower on Type 2 platforms because the host OS file system adds an extra translation layer.
Licensing Considerations
Licensing is an increasingly important factor in hypervisor selection. VMware's shift to subscription-only licensing under Broadcom has significantly increased costs for many organisations. Microsoft Hyper-V is included with Windows Server licences but requires per-core licensing. Proxmox VE and KVM are open-source and free to use, though production environments should budget for support subscriptions. When calculating total cost, consider not just the hypervisor licence but also management tools, backup integration, and the skills required — retraining staff for a new platform is a real and often underestimated cost.
KVM is technically a "Type 1.5" hypervisor. While KVM runs within the Linux kernel, the kernel itself runs directly on hardware, and KVM modules turn it into a hypervisor. This means KVM has the performance characteristics of a Type 1 hypervisor while benefiting from the vast Linux driver ecosystem. It is classified as Type 1 by most industry analysts despite running as a kernel module within Linux.
Frequently Asked Questions
While technically possible, it is strongly discouraged. Type 2 hypervisors add unnecessary overhead, lack enterprise management features, and depend on the stability of the host OS. If the host OS crashes or requires a reboot for updates, all VMs go down with it. Use a Type 1 hypervisor for any production workload.
Absolutely. Proxmox VE supports live migration, high availability clustering, ZFS storage, Ceph integration, and both VMs and LXC containers. For small-to-medium businesses, it offers enterprise-grade features at a fraction of the cost. The main trade-off is a smaller ecosystem of third-party integrations compared to VMware.
You need a CPU that supports hardware-assisted virtualisation (Intel VT-x or AMD-V), which has been standard on all server and most desktop CPUs for over a decade. For production use, you also want ample RAM (virtualisation is typically memory-bound), fast storage (SSDs are strongly recommended), and multiple network interfaces. Check your chosen hypervisor's hardware compatibility list before purchasing.
Containers (Docker, Kubernetes) and VMs serve different purposes and are complementary, not competing. Containers share the host kernel and are ideal for microservices and application packaging. VMs provide full OS isolation and are necessary when you need to run different operating systems, require kernel-level separation for security, or run legacy applications. Most modern environments use both.
