GPU Server User Guide

Instructions for setting up a remote development environment on the lab GPU server.

Prerequisites

  • PyCharm Pro (free with university email)
  • Personal server account (contact server administrator)

Step 1: Install Miniconda

Connect to the server via SSH:

ssh -p [PORT] [username]@[SERVER_IP]

Create and enter the downloads folder:

mkdir downloads
cd downloads
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh

When prompted "Do you wish the installer to initialize Miniconda3 by running conda init?" — answer yes. Then log out and reconnect.

Step 2: Create Conda Environment

conda update conda
conda create -n [ENV_NAME] python=[VERSION]
conda activate [ENV_NAME]

Install required packages for your project.

Step 3: Connect PyCharm to Remote Server

  1. Install PyCharm Pro
  2. Create a new project and add a remote interpreter
  3. Enter server IP, port, and account ID
  4. Set interpreter path: /home/[username]/miniconda3/envs/[ENV_NAME]/bin/python3

Step 4: Deploy Project Files

PyCharm automatically syncs files on save. Check remote path via Tools → Deployment → Browse Remote Host.