Back

vGPU (virtual Graphics Processing Unit)

A vGPU, or virtual Graphics Processing Unit, is a technology that allows the power of a physical GPU to be shared across multiple virtual machines (VMs). This is achieved by installing virtual GPU software on a physical GPU in a server, which then creates virtual GPUs that can be allocated to different VMs. This allows for efficient use of GPU resources, enabling multiple users to access GPU acceleration for tasks such as graphics-intensive applications, data science, and artificial intelligence workloads[1][3][5].

Citations:

[1] https://www.nvidia.com/en-us/data-center/virtual-solutions/

[2] https://www.datacenters.com/news/what-is-a-vcpu-and-how-do-you-calculate-vcpu-to-cpu

[3] https://blogs.nvidia.com/blog/what-is-a-virtual-gpu/

[4] https://www.techtarget.com/whatis/definition/virtual-CPU-vCPU

[5] https://en.wikipedia.org/wiki/GPU_virtualization

[6] https://phoenixnap.com/kb/what-is-a-vcpu

[7] https://krutavshah.github.io/GPU_Virtualization-Wiki/

[8] https://openmetal.io/docs/glossary/what-is-vcpu/

[9] https://documentation.suse.com/sles/15-SP3/html/SLES-all/article-nvidia-vgpu.html

[10] https://www.accuwebhosting.com/blog/what-is-virtual-processor-or-vcpu/

[11] https://youtube.com/watch?v=NiXtswuE1MI