- 16 CPU cores and
- 512GB of RAM, and
- 4x NVIDIA Tesla V100 GPUs interconnected with NVlink2.
- VRAM for each GPU is 16, totaling 16x4 VRAM
- Accessible from anywhere via SSH
To gain access to the High-Performance Computing (HPC) resources provided by Temple University, you can follow these steps:
- Visit the HPC website at https://www.hpc.temple.edu/.
- Choose the appropriate option based on your needs, such as machine learning, computing, or owl nest. Each option offers different features and criteria.
- Click on "Request Access."
- Specify your requirements, including your field, a description of your project, resource requirements, required software, and your supervisor's name.
- Read and agree to the terms and conditions.
- After submitting your request, your documents will be reviewed, and you should receive access within 2 to 3 business days.
-
Access via vscode:
- The best way I found to access HPC is from VS code.
- Install an extension called “Remote- SSH”
- Click on the Remote- SSH extension and write-
-
Access via Jupyter Nodebook:
- From the VSCode terminal create: python3 -m venv env (In the base environment, you don’t have any access to install any package. You need to create a new environment.)
- To install Anaconda write the command: :
wget https://repo.anaconda.com/archive/Anaconda3-2020.07-Linux-x86_64.sh
bash Anaconda3-2020.07-Linux-x86_64.sh
- To install jupyer notebook write the command (it will take time to fix some dependencies):
conda install jupyter notebook
- If you want to run the jupyter notebook write cmd:
- Copy the token (ex: http://127.0.0.1:8889/?token=22dd2ab63c90249bf4fa3da5ba468f5ceb034f64ff04) and paste in your browser. Here you go!!
-
On your Mac or Linux PC, open up your terminal and run the following command (unless you already have an .ssh/config file.)
touch ~/.ssh/config
-
you will see files in your .ssh folder. E.g: config, id_rsa, id_rsa.pub
-
Use nano or any text editor you prefer to copy and paste the following in the config file. Of course replace your User name with your AccessNet name.
Host Temple-GPU-4 # ssh [email protected] HostName 129.32.84.168 User tuq24452 Port 22
-
Create SSH Key in your local computer and add it to each server's id_rsa file.
- This cmd will create new key:
ssh-keygen
- Then press enter to save the key into id_rsa.pub
- This cmd will create new key:
-
Here we go. We don't need to put passward each time.