This tool is made to help you select the best vGPU profile based on GPU utilization of existing workstations. You run the tool for a while on your workstations and it records your GPU utilization and framebuffer utilization. It also looks at your screen resolution and number of monitors. Based on this information and feedback and advice from Erik Bohnhorst at NVIDIA, I have developed algorithms that convert the peak framebuffer utilization to a vGPU profile. I also look at the capacity of your existing GPU and try to convert this to the capacity of the NVIDIA GRID cards. With this information you can easily see what kind of hardware you will need to virtualize your workstations and how it will scale.
This is a sample workload, and you can quickly see that this workload could scale up to 128 users with an M10 card. You can also find out which applications are using DirectX, OpenGL, OpenCL, and CUDA.
In this video you can see how Autodesk 3DStudio requires a minimum 2GB profile on Windows 10, and that the K1 GPU is not recommended, because it isn’t powerful enough.
Performance data collected
- GPU Framebuffer utilization
- GPU load
Sponsor version also provides information about
- CPU type and utilization
- Memory utilization
The tool collects data about processes that has DLL’s loaded for GPU utilization. In this way you get a list of apps that can benefit from GPU. Supported API’s is:
- DirectX 9-12
One important thing with vGPU sizing is to know the number of monitors and the screen resolution on the client computer. GPUSizer will select a vGPU profile that is compatible with the number of monitors and the screen resolution that you have.
The tool works with all Nvidia Quadro and GRID cards if you have the latest drivers and NWMI installed. There is limited support for K1 and K2 vGPU because they do not provide performance data through API.
For other GPU types, it will estimate full load on the GPU and give recomendations based on this, this may give you oversized vGPU profiles, but it will still help you plan your environment. The reason for this is that there is limited GPU performance monitoring API’s available for non NVidia GPU’s. This will hopefully be possible in the future.
The tool is developed with .Net 4.5 and requieres this to work. It has been tested with Windows 7-10 and server 2008R2 – server 2016.