GPU Compute


The ever increasing demand for a GPU Compute facility by the IIT Guwahati reseach fraternity, led CCC to initiate deployment of this facility. This page covers a wide range of topics ranging from a detailed description of the hardware infrastructure to the information required to utilize the facility, such as information about logging on to the GPU Compute facility, submitting jobs, etc. In short, this page describes all that one needs to know to effectively utilize the GPU Compute facility.

The GPU Compute facility is based on a simple configuration of one login node and one GPU node. The system was designed and implemented by HPC IIT Guwahati Team in collaboration with Converge Systems and Services Pvt. Ltd.

Overview


System Hardware specifications of GPU Compute system is based on processor Intel(R) Xeon(R) 6740E 2400 Mhz and NVIDIA RTX PRO 6000 Blackwell DC with single-precision performance FP32: 120 TFLOPS

Total number of nodes: 2 (1 + 1)

  • Login nodes: 1
  • GPU nodes: 1

Login Node


Login nodes are typically used for administrative tasks such as editing, writing scripts, transferring files, managing jobs, and so on. You will always get connected to the login node. From the login node, you can get connected to the GPU node and execute an interactive job or submit batch jobs through the batch system (SLURM) to run your jobs on the GPU node. For all users the login node is the entry point. By default, there will be a limit on the CPU time that can be used on a login node by a user and there is a limit/user on the memory as well. If any of these are exceeded, the job will get terminated.

Specification:

  • Login Nodes: 1
  • CPU: Intel(R) Xeon(R) 6507P 3500 Mhz
  • Cores: 8 cores, 16 threads
  • Memory: 32 GB
  • Usable Storage: 446.63GiB for OS (2x 480GB 6.0G SATA SSD in RAID1)

GPU Node


The GPU node is the workhorse of GPU Compute facility. All the GPU intensive activities are carried out on this node. Users can access this node only from the login node to run interactive or batch jobs. The GPU compute node has CPU cores and GPU cards.

Specification:

  • GPU Nodes: 1
  • CPU: 2x Intel(R) Xeon(R) 6740E 2400 Mhz
  • Cores: 192 cores; 192 threads
  • Memory: 1536 GB
  • Usable Storage: 894.19GiB for OS (2x 960GB 16G NVMe SSD configured in RAID1); 20.96TB for /home, 20.96TB for /scratch, (8x 7.68TB 64.0G NVMe SSD configured in RAID5);
  • GPU Cards: 8x NVIDIA RTXPRO6000BlackwellDC - each equipped with 96 GB GDDR7 memory

Operating System


  • Operating System on GPU Compute is Linux – Alma Linux 10.x

Click Here to Know More About GPU Compute facility.

 

If you have login access issues, Click here for self help.

Or

Send an email to hpcadmin@iitg.ac.in