Ubuntu Server vGPU 16.8 24.04.1
Description¶
Ubuntu is a Linux distribution based on the Debian architecture. Ubuntu is the most popular operating system running in hosted environments. Ubuntu is the most popular Linux server distribution.
Ubuntu Server vGPU 16.8 24.04.1 LTS has not graphical user interface. This service is provided with Nvidia Tesla V100 32GB GPU.
Features¶
| Parameter | Value |
|---|---|
| Operating System | Ubuntu Server + Nvidia vGPU 16.8 |
| OS Release | 24.04.1 LTS (long-term support) |
| Desktop Environment | - |
| License | GPLv2 and other (Free and open-source) |
| Processor | 1 CPU |
| RAM | 1 GiB |
| Swap | 1 GiB |
| System Disk | 10 GiB |
| User Account | admin |
| User Password | Options: temporary password (sent by email); set own password |
| Network Interfaces | eth0 managed by Netplan |
| Initialization System | cloud-init |
| Virtualization | oVirt |
| Updates | 2024-11-06 |
Disk Partitioning¶
Partition Table: GPT
Filesystem / (ext4) auto-resizes to provisioned system disk size.
| Type | Name | Size, GiB | File System | Description |
|---|---|---|---|---|
| Partition | sda1 |
0.5 | vfat | EFI |
| Partition | sda2 |
1 | ext4 | boot |
| Partition | sda3 |
8.5 | - | LVM PV |
| LV | ubuntu-root |
7.5 | ext4 | root |
| LV | ubuntu-swap |
1 | swap | swap |
vGPU Information¶
| Parameter | Data |
|---|---|
| VM Installed Driver | NVIDIA-Linux-x86_64-535.216.01-grid.run |
| NVIDIA Driver Version | 535.216 |
| NVIDIA vGPU Software Version | 16.8 |
| Supported on VM CUDA Toolkit Version | 12.0 (not install) |
Please DO NOT UPDATE the Nvidia driver and DO NOT INSTALL the full CUDA as this will reinstall the driver and it will not work.
To install the CUDA Toolkit without the driver (more than 7 GiB), use the following commands:
To check the status of your Nvidia GPU and the installed CUDA version, use the following commands:To reinstall the driver run the command:
In the dialog box, select all suggested default values.Tested vGPU configurations¶
| Parameter | Data |
|---|---|
| Model GPU | Tesla V100 32 GiB |
| Video memory size | 32 GiB, 64 GiB |
| NVIDIA vGPU software version | 16.8 |
| Host NVIDIA driver version | 535.216 |
| VM NVIDIA driver version | 535.216 |
| Host CUDA version | 12.2 |
| VM CUDA Toolkit version | 12.0 |
Configuration¶
OS firewall is disabled. Use app Networking > SecurityGroups instead.
Installed software¶
- qemu-guest-agent
- cockpit
- ssh
- special driver
NVIDIA-Linux-x86_64-535.216.01-grid.run -dkmsand additional packages to it:build-essential,gcc-multilib,dkmslibvulkan1. where:build-essential- for building the driver,gcc-multilib- for providing 32-bit support (optional),dkms- for automatic rebuilding of the Nvidia GPU driver when updating the OS,libvulkan1- for supporting low-level access to the GPU using Vulkan ICD (optional).
License¶
Free and opensource: GPLv2 and other free software licenses. Paid licenses and subscriptions are not required.