Artificial Intelligence Computing Leadership from NVIDIA
Nvidia Grid Tesla P40 performance issue after vgpu driver installation on vdi
Hello NVIDIA Grid Forum, actually we are working on a project with cad virtualization. i have a big performance issue with my Citrix/Nvidia Grid Infrastructure the problem starts after we install the nvidia GPU drivers in the VMs/VDIs. different applications are very slow after the driver installation for example solidworks(CAD Application) oder a Sage Client(ERP Application) Before the installation everything is OK!!!!! this is our infrastructure: 2 x Dell PowerEdge Server 740 each server has two NVIDIA Tesla P40 Grid cards we tested it on Xenserver 7.1 / 8.0 / 8.1 full patched we tested it with vGPU 9.2 (431.79) und 10.1 (442.06) the problem exists with different nvidia profiles 2Q or 12Q the VMs working with Win10 1903, the problem is also there with Win10 1909 the problem exists with 32GB RAM or if you assign a lot of more RAM the problem is also there with only one vm on the server at the same time you can see the problem directly in the console or in a citrix session(same impact) we tried a lot of things but at the moment the ideas run out ;-( Do somebody have ideas for us or the same impact? thanks a lot Zoran
Hello NVIDIA Grid Forum,

actually we are working on a project with cad virtualization.
i have a big performance issue with my Citrix/Nvidia Grid Infrastructure
the problem starts after we install the nvidia GPU drivers in the VMs/VDIs.
different applications are very slow after the driver installation
for example solidworks(CAD Application) oder a Sage Client(ERP Application)

Before the installation everything is OK!!!!!

this is our infrastructure:
2 x Dell PowerEdge Server 740
each server has two NVIDIA Tesla P40 Grid cards
we tested it on Xenserver 7.1 / 8.0 / 8.1 full patched
we tested it with vGPU 9.2 (431.79) und 10.1 (442.06)
the problem exists with different nvidia profiles 2Q or 12Q
the VMs working with Win10 1903, the problem is also there with Win10 1909
the problem exists with 32GB RAM or if you assign a lot of more RAM
the problem is also there with only one vm on the server at the same time
you can see the problem directly in the console or in a citrix session(same impact)


we tried a lot of things but at the moment the ideas run out ;-(

Do somebody have ideas for us or the same impact?

thanks a lot


Zoran

#1
Posted 03/14/2020 11:09 AM   
Scroll To Top

Add Reply