Gihan Chanaka Jayatilaka : Resources : Peradeniya HPC Documentation

There is an Official FAQ for CO servers. This document contains some documentation that I wrote, which is yet to be merged to the official FAQ. In addition, there is a similar documentation for EE servers as well.

If you want to do DL on any of the servers, first create a conda environment and install pytorch/tensorflow on it. You can run the code through SSH or Jupyter notebooks after that.

• using-conda-without-sudo-peradeniya-servers: Merged to CO FAQ here
• pytorch-tensorflow-versions-for-Peradeniya-servers
• Jupyter notebooks: Merged to CO FAQ here.
• VS code SSH development for Turing/Kepler. Then, you can install pylance for the Turing server and select the correct conda environment for code completion.
• Requesting/managing project storage on Peradeniya Servers: Merged to CO FAQ here.
• Using Mobaxterm to access the terminal and file storage on Turing/Kepler. Here, the public IP server (Tesla/Aiken) is the jump host and the internal IP server (Kepler/Turing) is the remote host.
• Minimizing the storage usage on Turing localhomes (We no longer use turing:/localhomes)
• FAQ for new HP Server (2023): Merged to CO FAQ here.
• Server usage statistics (ie. How busy are the servers?)
• Server groups, folders and datsets (These pages should be accessible to anyone with @eng.pdn email gsuite accounts): Turing, Kepler, Aiken, New HP Server (2023).
• Sudoers on Turing/Keple/Aiken/Tesla?: sheet
• Managing storage on the babbage server
• Singularity containers: Ask for help from Kalindu or Sashini

• COVIDAI-GPU-Server. This file should be accessible though EE Server Admin.
• What GPU to buy next for Peradeniya?

• We are looking for someone to copy the content on this webpage to faq.ce.pdn.ac.lk. If you are interested, please contact Akila/Nuwan via email.
• We are looking for students to work on some projects to improve the ML/DL tools for Peradeniya servers. [Read more]