QEMU vs KVM in Proxmox: What Is the Difference?

QEMU and KVM are frequently discussed as competing virtualization technologies, but that comparison is misleading. In Proxmox, they perform different parts of the same virtual machine process.

KVM provides hardware accelerated processor and memory virtualization through the Linux kernel. QEMU creates the virtual computer around the guest, including its disks, network adapters, firmware and other devices. Proxmox adds the management platform that lets administrators configure and control the complete environment.

A Proxmox virtual machine normally uses QEMU and KVM together. Understanding their roles helps explain virtual machine performance, hardware compatibility, device selection and troubleshooting.

What Is QEMU?

QEMU is an open source machine emulator and virtualizer. Its system emulation creates a virtual model of a computer, including processors, memory and hardware devices, so that a guest operating system can run as though it were installed on physical hardware. citeturn281323search1turn285123search15

QEMU can model many processor architectures and machine types. It can emulate x86, Arm, RISC V and several other architectures. This allows QEMU to run software designed for a different processor architecture from the physical host, although full emulation is much slower than hardware assisted virtualization. citeturn281323search23turn281323search39

From the Proxmox host’s perspective, each QEMU virtual machine runs as a user space process. That process represents the virtual hardware presented to the guest operating system.

QEMU can provide virtual disk controllers, network cards, USB controllers, graphics adapters, serial ports and many other devices. A Windows or Linux guest sees those components as hardware and loads drivers for them in the same way it would on a physical computer. citeturn281323search10turn285123search29

QEMU can operate without KVM by using its Tiny Code Generator. TCG translates guest processor instructions into instructions the host can execute. This makes cross architecture emulation possible, but it is usually too slow for ordinary production server workloads. citeturn285123search1turn281323search30

What Is KVM?

KVM stands for Kernel based Virtual Machine. It is a virtualization subsystem inside the Linux kernel.

KVM uses processor virtualization extensions, such as Intel VT x or AMD V, to allow guest processor instructions to execute efficiently on the physical processor. The Linux kernel documentation describes KVM through an API used to create virtual machines, configure virtual processors and control virtual machine capabilities. citeturn281323search3turn281323search43

KVM does not provide a complete graphical management system, storage platform or collection of virtual devices by itself. It provides the kernel level virtualization foundation.

The Linux kernel handles scheduling, memory management and access to the physical processor. KVM adds the ability to create isolated virtual processor and memory environments for guest systems.

This is why KVM is sometimes described as turning the Linux kernel into a hypervisor. That description is broadly useful, although a functioning virtual machine still needs a user space program such as QEMU to create and manage the rest of the virtual computer.

How Do QEMU and KVM Work Together?

When Proxmox starts a virtual machine, QEMU creates the virtual machine process and its virtual hardware. QEMU then uses KVM as an accelerator so that most guest processor instructions can run directly through the host processor’s virtualization extensions.

QEMU remains responsible for the virtual machine model. KVM accelerates the processor intensive parts of running the guest.

A simplified sequence looks like this:

Proxmox reads the virtual machine configuration.

Proxmox starts the appropriate QEMU process.

QEMU allocates the virtual memory and creates the configured devices.

QEMU connects to the Linux KVM interface.

KVM runs the guest virtual processors using hardware virtualization.

QEMU handles virtual disks, network devices and management events.

The Proxmox administration guide explains that, in the Proxmox context, QEMU and KVM are often used interchangeably because Proxmox QEMU virtual machines normally attempt to use the Linux KVM module. citeturn285123search29

This does not mean they are the same software. It means administrators rarely use one without the other when running normal Proxmox virtual machines.

Where Does Proxmox Fit?

Proxmox VE is the management layer above QEMU and KVM. It integrates KVM based virtual machines and LXC system containers into a platform with a web interface, command line tools and an application programming interface. citeturn281323search9turn281323search13

Proxmox manages virtual machine configuration, storage, networking, backups, snapshots, clustering, migration, permissions and high availability.

When an administrator selects four virtual processors, 16 gigabytes of memory and a VirtIO network adapter in the Proxmox interface, Proxmox converts those choices into the configuration used to start and control the QEMU process.

The qm command line tool is Proxmox’s QEMU and KVM virtual machine manager. The virtual machine configuration is stored by Proxmox, while QEMU and KVM perform the actual virtualization work. citeturn285123search31turn285123search20

Proxmox therefore does not replace QEMU or KVM. It makes them easier to operate across individual servers and clustered infrastructure.

What Does QEMU Emulate?

QEMU can emulate hardware that existing operating systems already understand. This is useful when installing an older operating system or importing a virtual machine that does not yet contain optimized virtualization drivers.

For example, QEMU can present familiar IDE, SATA or emulated network controllers. The guest may boot immediately because drivers for those devices are already included in the operating system.

The disadvantage is performance. Emulating physical hardware in software requires additional processing because every device interaction must be translated.

Proxmox therefore also supports paravirtualized VirtIO devices. With VirtIO, the guest understands that it is running in a virtualized environment and communicates with the host through optimized interfaces rather than fully emulated hardware. Proxmox recommends VirtIO SCSI for many migrated virtual machines because it provides efficient disk access and supports features such as input and output threads. citeturn285123search2turn285123search29

Windows guests usually need additional VirtIO drivers. Linux distributions commonly include them already.

Does KVM Handle Virtual Disks and Networks?

KVM does not independently create the complete virtual disk or network device visible to the guest.

QEMU presents the virtual controller and connects it to a storage resource or network backend. Proxmox decides where the disk image is stored and which bridge or virtual network the interface uses.

KVM may accelerate certain parts of device handling through kernel components, but QEMU remains the virtual machine monitor coordinating the device model.

This separation explains why a QEMU related error may involve a disk, virtual network adapter or passed through PCI device, while a KVM related error may point toward processor virtualization, kernel modules or unsupported CPU features.

Does QEMU Reduce Proxmox Performance?

QEMU itself does not mean that every processor instruction is being slowly emulated.

When KVM acceleration is active, guest processor instructions can run through hardware virtualization. QEMU mainly provides the surrounding virtual machine and device environment.

Performance differences are more likely to come from the selected virtual hardware, storage design, processor model and workload.

Using emulated disk or network devices may create more overhead than VirtIO devices. Selecting an overly generic virtual processor can hide useful host processor features. Selecting the host processor type can expose more features, but it may restrict live migration to nodes with compatible hardware.

Device passthrough can reduce emulation overhead further by assigning a physical device directly to the guest. QEMU officially supports direct device assignment when the host platform provides the required capabilities. citeturn285123search1

Can Proxmox Run Without KVM?

QEMU can technically run without KVM by using software emulation, but this is not the normal Proxmox production model.

A Proxmox server should have hardware virtualization enabled in its firmware. If Intel VT x or AMD V is disabled, unavailable or already controlled by another virtualization layer, KVM acceleration may fail.

Nested Proxmox installations can use KVM when the outer hypervisor exposes nested virtualization extensions. The Linux KVM documentation includes support for nested Intel virtualization, but the complete environment must support and expose the required processor capabilities. citeturn281323search11

Without KVM acceleration, virtual machine performance can be drastically lower. This is why confirming hardware virtualization support is one of the first troubleshooting steps when building a Proxmox host.

Is QEMU or KVM More Important?

Neither component is sufficient alone for the normal Proxmox virtual machine experience.

KVM provides efficient access to the processor’s virtualization capabilities. QEMU creates the virtual hardware and manages the user space virtual machine process. Proxmox turns those technologies into an operational platform with storage, networking, clustering, backups and centralized management.

Calling a Proxmox guest a KVM virtual machine emphasizes the hardware accelerated Linux virtualization layer. Calling it a QEMU virtual machine emphasizes the process and virtual hardware model.

Both descriptions refer to the same practical workload.

The clearest way to understand the stack is simple: KVM accelerates the virtual processor, QEMU builds and operates the virtual computer, and Proxmox manages the infrastructure around it.