Vipul Rathore

Compute servers

  • There are 3 GPU compute nodes -
    1. Shakuntala1 (3 GTX1080, 40 cores) (Note:- ssh to shakuntala1 should be done using the IP Address 10.237.23.117 (ssh user@10.237.23.117))
    2. Shakuntala2 (6 GTX 1080, 64 cores) (ssh user@shakuntala2.cse.iitd.ac.in)
    3. Ramanuja1 (1 V100, 40 cores) (ssh user@ramanuja1.cse.iitd.ac.in)
  • We recommend users to use Anaconda for installing packages. Use the required cudatoolkit (if needed) in your conda environments. You can create multiple environments.
    Note:- WE PREFER NOT TO update the existing CUDA drivers on these systems. It may cause problems for other users using the same system.
  • If you are using the multi-GPU servers (shakuntalaX), then please use them responsibly. Don’t hog all the GPUs unnecessarily. We recommend maximum 2 nodes on shakuntala2 per user at any time. If you need more, you should coordinate amongst yourselves on the WhatsApp group.
  • To connect to the internet from your terminal, use the proxy scripts. More details can be found here (Courtesy: Anupam Sobti). This is an internal link. Using screen sessions is a good idea for working remotely.
  • Since the disk space is limited, each user will get 200 GB space in /home by default. To make best use of it, please take regular backup of your data in local machines/hpc/onedrive or any other persistent storage of your choice. You can request up to 2TB free home space in HPC through your advisor. Onedrive provides 1 TB free space for IITD users.
    NOTE: shakuntala/ramanuja servers are for purely compute purpose and not for storage. Your files and folders in /home are liable to be deleted if the disk space starts to get full. We will try to delete the older files on a best effort basis (no-guarantees).
  • DO NOT download large datasets/models from internet directly in the shakuntala/ramanuja servers. Download it in your local/HPC first and then transfer to the shakuntala/ramanuja servers. You may consider using the research proxy for bulk download.
  • Multi-GPU server (shakuntalaX) server users may use the designated WhatsApp channel for inter-user communication and raising issues to the admin.

    • Storage servers

  • There are 3 CPU storage servers -
    1. Panini1 (40 cores) (ssh user@panini1.cse.iitd.ac.in)
    2. Panini2 (40 cores) (ssh user@panini2.cse.iitd.ac.in)
    3. Analytics2 disk (100 TB) (/misc/research/analytics2 which can be accessed from sri/panini1/panini2)
  • All the above storage servers are linked with GCL account. You need to create an account on GCL (visit GCL lab in Bharti building, 4th floor and submit a form to lab attendant) as a first step to access these servers.