Hi
I am new to using the Nvidia Tesla M2075 on ESXi 5.0
Has anybody know how to enable the GPU or provide a guide so that I can enable the GPU on the Dell R720 Machine.
I am not seeing the card on my vSphere Client Conifguration Page.
I am using Dell customize ESXi CD.
Thks for the advice in advice
You could try to use DirectPath I/O and dedicate the card to a specific virtual machine.
// Linjo
Could you give us some more details on what the usecase is?
// Linjo
Basically, I am trying out something different, to run image rendering software on the server.
You could try to use DirectPath I/O and dedicate the card to a specific virtual machine.
// Linjo
Hi
sorry I had been trying to look for solutions, but it seems that my server does not detect the graphics card at all.
Kindly advice me how to set the driect I/O path
Rgds
Hi
Thks, after looking at the hardware list and finally got the correct passthrough, and now able to detect and I test my software, Thks
Rgds
Hi,
I am trying the same with Tesla 2090. Kindly let me know how u achieved GPU bypass. And how many VM's ur able to run concurrently
Hi
Sorry for the late reply, at the moment as I known of, if you are running windows VM Solution, you can, as the driver are detectable in windows 2008.
As for VM, passthrough works and can be detected, but it will not work in a full solution as the system cannot detect the GPU as shown in the image, you may want to wait for VM to release official driver for the Nvidia Telsa in future.
Hope that info help u abit.
Rgds
Thank u bubby thanks a lot for ur info
it will be great if u answer my further questions
Thanks in advance
Hi
Basically using the current version is 0.
I couldn't get anything to run.
As for the new release for ESXi 5.1, you may want to go for it as it support Tesla.
Detail I don;t have, may need to wait for official document from VMWare and Nvidia.
What are the minimum hardware and VMWare View required to run.
That all I have for now.
I don't know if this is a ridiculous concept, but has there been any research or talk regarding the use of Tesla GPUs as VM cores? Basically, I am talking about expanding the capacity of a host to support more VMs that rely on Tesla core processing.