HPC & DGX-A100

NVIDIA DGX A100 “First in India”

Over 20,000 TFLOPS total compute power

Over 2500 CPU Core

20 node High-Performance Computing (HPC)

400 TB + Storage

Highest grade AI Computing Infrastructure


The new 42-compute node with a total of 528 cores and high-speed Infiniband network with a peak performance of 6 TF High Performance Computing (HPC) cluster facilitates high-end computational research in the Institute covering many areas of science and engineering. This facility is specifically used for the following purposes:

a. High Performance Computing with several software.

b. Parallel programming under Linux using CUDA.

c. Parallel computing with C, C++, Fortran and Intel compiler

The technical details of this HPC facility are:

1. 5 Master nodes

2. 42 Compute nodes

3. 2 Compute node with tesla

4. 2 NFS server

5. 2 Backup Server

6. 2 IO server

7. 2 Parallel File System Server

8. 80TB Storage

9. 5 Infiniband Switch

10. 1 SAN switch

11. 1 Gigabit Switch

HPC Training & User Guide is available for its users.