IoV

IOV, short for I/O virtualization, refers to a set of techniques that allow input/output resources such as network adapters, storage controllers, or GPUs to be shared or partitioned among multiple operating contexts, typically virtual machines. The aim is to improve efficiency, reduce overhead, and enable scalable multi-tenant environments while preserving isolation between tenants.

The most widely deployed form is Single Root I/O Virtualization (SR-IOV), defined by PCI-SIG. In SR-IOV, a physical device presents a number of logical functions: a Physical Function (PF) used to manage the device, and one or more Virtual Functions (VFs) that can be assigned directly to virtual machines or containers. The host uses an IOMMU or similar mechanism to map memory addresses and provide device isolation. The result is that a VM can access a dedicated VF with near-native performance, while the remaining resources reside under host control.

Operation typically involves the virtualization software configuring the PF and VFs, exposing VFs to guest systems, and providing drivers inside the guests to communicate with the assigned VF. While VFs offer high performance and low overhead, some advanced features of the physical device may be unavailable inside VMs, and certain operations may complicate live migration or resource management across hosts.

IOV is widely used in data centers, cloud computing, and high-performance computing to improve NIC throughput, reduce CPU overhead, and enable finer-grained resource allocation. Limitations include hardware support requirements, compatibility of device features with VFs, and the complexity of managing virtualized I/O across dynamic workloads.