The Chandra High performance computing cluster (HPC) provides a powerful computing platform for research in engineering and physical sciences. This system has been operational since June 2017. The HPC consists of 64 compute nodes, each with a dual 12-core Intel processor. Each core runs at 2.2 GHz and has 4 GB of RAM per core. The HPC is one of the first systems in India to use a 100 Gbps high-speed OmniPath interconnect from Intel. The system provides about 50 TFlops of computing power. Chandra also accesses 100 TB of disk space setup as a parallel file system running Lustre from Intel.
The HPC is used by faculty, research staff and students at IIT Palakkad to investigate complex research problems in science and engineering. Some of the problems currently being studied are :
-
Understanding and designing materials with novel physical properties by performing atomistic quantum mechanical simulations.
-
Design of nanoscale transistors for next generation electronic applications
-
Design of novel bio-molecules with applications in medicine.
-
Design of large structures such as bridges and buildings.
-
Performing computational fluid dynamic simulations
-
Understanding the process of heat transfer in complex systems such as engines.
-
Solving non-equilibrium dynamics in quantum Hamiltonians
Hardware Specifications | |
Peak computing power | 54TFlops |
Total number of nodes | 64 |
Number of cores per node | 24 |
Total number of core | 1536 |
RAM per core | 4 GB |
Total RAM | 6 TB |
Storage Usable | 100 TB |
Interconnection Speed | 100 Gbps |