Logo
 
Department Site

Kepler

This is a GPU server of the Department of Computer Engineering, University of Peradeniya.

You can use this server using the LDAP login for ce.pdn.ac.lk undergraduates (which is used for CO top floor lab, Aiken and Tesla) without sudo access.

Since kepler.ce.pdn.ac.lk does not have a public IP, you can either SSH directly from inside the university network, or SSH through Tesla (which has a public IP) from outside the university network. Read how to SSH

Direct (from inside the university network): 
ssh e14000@kepler.ce.pdn.ac.lk

Jump through Tesla (from outside): 
ssh -J e14000@tesla.ce.pdn.ac.lk e14000@kepler.ce.pdn.ac.lk

  • CPU: Intel(R) Xeon(R) CPU E5-1630 v3 @ 3.70GHz | Cores: 8
  • RAM: 16 GB DDR4-2133 MT/s
  • GPU: Nvidia Tesla K40c and Nvidia Quadro K620

Ubuntu 20.4

  • Ubuntu 20.04.1 LTS
  • NVIDIA-Driver Version: 455
  • CUDA Version: 11.1 Update 1
  • cuDNN 8.0.4
  • conda 4.8.5 (UPDATE: You can install Conda by yourself)
  • cmake 3.16.3
  • gcc 9.3.0
  • openjdk 11.0.8
  • ffmpeg 4.2.4
  • SWIG 4.0.1
  • R 3.6.3
  • Blender 2.82

We store frequently used datasets inside the kepler.ce.pdn.ac.lk local storage for easy access. You can access these by the following command: 

ls /storage/datasets/

These datasets are in the read only mode. You can use these datasets as the input for your tasks but you have to output the results to your home directory. The following datasets are stored at the moment.

If you have suggestions for what datasets should be made available on this server, please send a request via email available in https://www.cs.umd.edu/~gihan/contact/ address.


You can access these by the following commands: 
ls /storage/datasets

You can try with following command: 
nvcc code.cu -arch=sm_35

The -arch=sm_35 flag is required to compile code properly on this server.


NVIDIA Tesla K40c is a Kepler GPU (compute capability 3.5). In practice, this limits support to the CUDA 11.x line. NVIDIA deprecated Kepler support in CUDA 11.4 and removed it entirely in CUDA 12.0.

A recommended PyTorch setup for this GPU is PyTorch 1.12.1 with torchvision 0.13.1, torchaudio 0.12.1, and CUDA 11.3: 

conda install pytorch==1.12.1 torchvision==0.13.1 torchaudio==0.12.1 cudatoolkit=11.3 -c pytorch

Please request through any CO staff member . Please note that the server admin has no authority to install something unless it is requested by a staff member.

Contact server admin

  1. Send a request to the server admin via the server admin contact page to run this command: 
    sudo singularity config fakeroot --add eXXYYY
  2. Download the sample def file: 
    eyyxxx@kepler:~$ wget https://raw.githubusercontent.com/cepdnaclk/singularity-deployment/main/base.def
  3. Build the container: 
    eyyxxx@kepler:~$ singularity build --fakeroot base.sif base.def
  4. Run commands using the container: 
    eyyxxx@kepler:~$ singularity exec base.sif <your-command>

If you need to learn more, go to: https://github.com/cepdnaclk/singularity-deployment


Ask via email