
## Prerequesites

Read [Connecting to a compute server](employee/compute-servers/connect) first.

## GPU

Two servers are equipped with a NVIDIA GPU: `r21` (_A30_)  and `r30` (_A40_).


You can check the GPU status with the `nvidia-smi` command:
```sh
ckohler@r21-vm:/home/ckohler $ nvidia-smi
```


If this returns
`Failed to initialize NVML: Unknown Error`
the container needs to be restarted due to a but related to cgroups and the nvidia driver.

Please contact [Michael Huttner](https://www.spang-lab.de/people/mhuttner) or [Christian Kohler](https://www.spang-lab.de/people/ckohler) for further assistance.

### Compiler

If you want to use the nvcc compiler you will need to add

`/usr/local/cuda/bin` to your `$PATH` in `~/.zshrc`.

```txt
...
export PATH=$HOME/.local/bin:/usr/local/cuda/bin:$PATH
...
```


GPU Servers

Get an acrux account

How to access the gitlab server

Spang Lab Services

SSH Keys

Jupyter

Business Trips and Travel Expenses

Deploying Software

How to Print

How to Print Posters

How to write articles

Mailinglists

Microsoft Teams@UR

Vacation

Connecting to a compute server

Server Usage

Shared Publications

Articles that need to be written

Kubernetes Cluster

General Articles

Student Articles

Employee Articles

Compute Servers

PhD Articles

Sysadmin Articles

GPU Servers