I have been looking at header files, NVAPI, CUDA, Halide but I am unable to find the function to change the GPU Utilization from “Use for Graphics and compute needs” to “Dedicate to graphics tasks”. Looks like it is something to do with Maxwell Technology but I am unable to find an API. Any help would be greatly appreciated Thank you so much
The NvCpl example with the NvPanelApi.h file is very old 2005 and I assume that the Workstation “Manage GPU Utilization” settings were not available back then.
I have tried using NVML COMPUTEMODE GOM_COMPUTE and they realized that they do not work with the NVS and Quadro video cards.
Also Tried Halide and CUDA but the function that makes the GPU Utilization change is not obvious in any of the header files and documentation I have read. The setting it does appear is based on Maximum Technology?
When the change is applied it appears that NVCPL.DLL is calling a function in NvDispS.dll for which I can find no information.
Thank you so much
Any update on this NVIDIA?
I have the same issue with deployments in my environment and currently must manually change this setting on each workstation.
Is there a valid reason for settings available in the NVIDIA Control Panel to NOT be changeable through NVWMI?
If not, is there an anticipated date that this setting will be added?
Is there another command line solution for changing this setting either at time of scripted driver deployment or through Group Policy (script or registry setting)?