An organization purchased new servers with GPUs for render farms. The servers have limited CPU resources. Which of the following GPU configurations will be the MOST optimal for virtualizing this…

Dedicated GPU assigns an entire physical GPU to a single VM, which wastes GPU resources when multiple render VMs need simultaneous access.

Shared GPU typically refers to a software-based time-slicing approach with no resource guarantees, resulting in contention and inconsistent render performance.

GPU passthrough (PCIe passthrough) grants exclusive ownership of a physical GPU to one VM, making it unavailable to other VMs and unsuitable for a multi-VM render farm.

vGPU technology (such as NVIDIA vGPU) partitions a physical GPU into multiple virtual GPU instances, each assigned to a separate VM with guaranteed compute and memory resources. This allows many render VMs to use GPU acceleration concurrently without requiring a dedicated physical GPU per VM. Because CPU resources are limited, offloading rendering to shared GPU slices maximizes hardware utilization across the virtualized environment.