I’ve installed server core 2016 and hyper-v on HPE hardware, DL380 G10 and passthrough the P4000 card successfully but in device manager I receive error code 43.
The VM OS is server 2012R2 with the driver 391.33-quadro-tesla-grid-winserv2008-2008r2-2012-64bit-international-whql
What’s weird is that we did this exact same setup in April this year and that deployment works flawlessly, how can we solve this? Is DDA supported by Nvidia?
not sure what the issue is here but in general that should work. The main issue I see here is core server. This is not tested from NV and therefore might cause the error 43. DDA with Quadro Passthrough should work without issues.
The issue is that the graphics card seems to not work properly. I might have confused you, the OS hosting the VM is server 2016 core installation and I’ve given the 2012R2 VM on this host the graphics card but after installing the driver and after reboot the card returns code 43 but the correct name is shown in device manager.
I tested the P4000 in a baremetal deployment (no hypervisor involved) and that installation worked flawlessly so the card is working and the drivers seems to be working as well.
Windows server 2016 Core installation with hyper-v role installd. The hardware is HPE DL380 with Xeon Gold 6132 CPU.
This server hosts one virtual machine which runs Windows server 2012R2 and I’ve successfully passthrough the graphics card and after installing the driver it shows up after reboot as Nvidia Quadro P4000 but with the message "Windows has stopped this device because it has reported problems. (Code 43)".
Does anyone or Nvidia know what kind of problem we’re facing? When I google code 43, all results are incompatible graphic cards and this does not seems to be the case for me as P4000 should be compatible iun virtual machines, just like Simon describes.
We’re very eager to make this work but if we can’t get it to work with DDA, we’ll have to look at other options.
I assume something DDA specific is still missing. Did you do the following:
snip
Some Additional VM preparation is required for Graphics Devices
Some hardware performs better if the VM in configured in a certain way. For details on whether or not you need the following configurations for your hardware, please reach out to the hardware vendor. Additional details can be found on Plan for Deploying Devices using Discrete Device Assignment and on this blog post.
Enable Write-Combining on the CPU Set-VM -GuestControlledCacheTypes $true -VMName VMName
Configure the 32 bit MMIO space Set-VM -LowMemoryMappedIoSpace 3Gb -VMName VMName
Configure greater than 32 bit MMIO space Set-VM -HighMemoryMappedIoSpace 33280Mb -VMName VMName Note, the MMIO space values above are reasonable values to set for experimenting with a single GPU. If after starting the VM, the device is reporting an error relating to not enough resources, you’ll likely need to modify these values. Also, if you’re assigning multiple GPUs, you’ll need to increase these values as well. snip
Thank you Simon for your help, unfortunately, this didn’t help my case but I agree that this seems to be more of a DDA problem rather than Nvidia driver or the graphics card as a whole.
Thank you for your quick responses and if I solve this with Microsoft I’ll post my solution for others to see who might encounter the same issue.
I face a similar issue. I use a Nvidia Quadro P4000 on a HPE DL380 Gen10 with Windows Server 2016 Hyper-V and Passtrough the Card with DDA to a Windows Server 2016 VM.
After installing the latest driver, I also got the error code 43.
Microsoft claimed the driver whose at fault and that the resolution was to activate nested virtualization and passthrough the CPU but this didn’t work either, still code 43. Nested virtualization is not an option for us to use but I had to test if it would work.
The technician at Microsoft didn’t see any errors in my setup. DDA passes through the graphics card without trouble and like I said before, they claim that it’s the driver which shut downs the graphics card.