NCHC-CT

This tutorial goes through the usage of NCHC container for writing CUDA programs. The system file will be reset on every launch, but the files located at /root/ will be saved and therefore you should save all of your code under /root. Nonetheless, you should always backup your code such as using git.

Website: https://portal.educloud.nchc.org.tw

Getting started

Please follow the tutorial prepared by previous TA to start the container and obtain the ssh port for accessing the container: https://hackmd.io/ynRIWGL1Rvytr4Z-uvuRfw

TL;DR:

1. You can select using 1 or 2 GPUs by starting the corrosponding container.
   

   

2. Next, navigate to "工作清單". The container should appear in this page.
   

   

3. Access your container. We provide 2 options:

   • SSH:
     Click the "ssh" button. Your browser will open an invalid webpage. The link should be "nodeport.educloud.nchc.org.tw:xxxxxx". You can access the container by
     ssh [e-mail] -p xxxxxx, where the password is student.
     
     
   • Visual Studio Code Server:
     We provide an easier way to access your container. This is a webpage acting as a normal visual code app. To access the webpage, click "vscode-server" and login with the password student.
     
     

4. Install the one-time script.

One-time script

This script prepares the following 2 settings:

1. Generate SSH key
2. Copy SSH key to apollo server. This step will make the judger recognizes your username, instead of root.

During running the script, you will be ask to enter your username and password for logging in to Apollo. Since the files under /root/ will be saved, this script only requires running once.

bash <(curl -s https://apollo.cs.nthu.edu.tw/ipc23/setup-remote.sh)

Writing your assignment

All sample files are located at /tmp/dataset-nthu-ipc23/share/. Once you complete the coding, execute hw3-judge to test the correctness. Results will be sent back to https://apollo.cs.nthu.edu.tw/ipc23/scoreboard/hw3. The container provides all tools needed by CUDA labs and assignments:

1. g++ (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0
2. CUDA 10.2
3. nvcc
4. nvprof & nsight compute (profiler)
5. GeForce GTX 1080 Ti * 2 (11 GiB VRAM)

PS:

• The default environment variables are configured for bash (~/.bashrc and ~/.bash_profile) only.
• The judger on NCHC will execute your program locally. Please avoid running any task during judgging.
• Mining is prohibited.