Loading...

ABOUT

Information



HPC at IIT Mandi

Welcome to the website of IIT Mandi supercomputing facility. The HPC supercomputing facility was established in 2012 at IIT Mandi campus. In the beginning, the facility consisted only of a few compute nodes with a total peak performance of 35 TFLOPS. Over the past few years, the facility has undergone rapid expansion and upgradation and the total peak performance is currently at 80 TFLOPS.
In addition, the institute houses the 838 TFLOPS PARAM Himalaya supercomputing facility commissioned under the National Supercomputing Mission (NSM). We also have several high-end CPU and GPU workstations to cater to the computational needs of the institute. All of our computational facilities are open to external users as well on a chargeable basis. If you are interested in using these facilities, please submit a formal application to us. We hope you find the website informative and useful.

Facility

Overview

PARAM

The Param Himalaya HPC Facility at IIT Mandi provides a powerful computing platform with a mix of CPU nodes, high-memory nodes, and GPU-accelerated nodes for a wide range of computational tasks. All nodes are interconnected via a High-Speed InfiniBand (IB) network, ensuring fast and efficient data transfer. Managed by SLURM, the cluster enables users to submit, monitor, and manage jobs with commands like srun, squeue, scancel, and sbatch. The facility also supports the use of pre-configured module files for efficient software environment management, with commands such as module load and module unload for easy software handling.



HPC

The HPC Facility at IIT Mandi is a multi-purpose Linux cluster designed for heavy computational tasks. It features 171 compute nodes equipped with Xeon Gold/Bronze processors, ranging from 64 to 128 GB of RAM, 10G connectivity, and 8 to 24 threads per node. The facility has two sub-clusters: CPUHPC, optimized for CPU-parallel jobs, and GPUHPC, which contains high-end Nvidia GPUs for GPU-intensive tasks. Available to all IIT Mandi members, the facility supports research and teaching. Access is granted via an HPC account, which can be requested through the official IIT Mandi HPC website.

Powerful Computational Infrastructure (Param Himalaya)

The Param Himalaya HPC Facility at IIT Mandi offers a robust computing environment with a combination of CPU nodes, high-memory nodes, and GPU-accelerated nodes. This diverse setup caters to a wide range of computational tasks, making it ideal for research, simulations, and large-scale data analysis.

High-Speed Network and Job Management (Param Himalaya)

The facility’s nodes are interconnected by a High-Speed InfiniBand (IB) network, ensuring low-latency, high-throughput data transfer. Managed by SLURM, users can efficiently submit, monitor, and manage jobs through commands like srun, squeue, and sbatch, streamlining the workflow for large computational tasks at IIT Mandi.

Hot Aisle Containment for Energy Efficiency (HPC Facility)

The HPC Facility incorporates hot aisle containment, optimizing air flow and cooling efficiency. This setup reduces energy consumption, ensuring the facility runs sustainably while maintaining optimal temperatures for the powerful servers. It contributes to both cost savings and improved performance of computational resources.

Scalable Compute Resources for Research (HPC Facility)

The HPC Facility at IIT Mandi features Xeon processors with 64 to 128 GB RAM, and 10G connectivity across 171 nodes. This scalable architecture supports a variety of research needs, from CPU-parallel tasks to GPU-intensive simulations, providing IIT Mandi's academic community with powerful resources for advanced computing tasks.

Our Team

HPC Team Members

Dr. Prasanth P Jose

Associate Professor

Dr. Bhaskar Mondal

Associate Professor

Anil Kumar

JTS

Kushal Bhatia

Project Associate

Jatin Sharma

Project Associate