High Performance Computing
SankhyaSutra Labs hosts SankhyaSutra Rudra – a petascale supercomputing cluster which is currently the 6th largest cluster in India.
The HPE Cray CS500 system has 256 nodes:
- Configured with two AMD EPYC 7742 processors per node
- Containing 512 Gigabytes of Samsung DDR4 3200 MT/s memory per node
- Interconnected with high speed Nvidia Mellanox HDR InfiniBand
- Accessing 6 Petabytes of HPE Cray ClusterStor global SATA HDD storage
SankhyaSutra Rudra’s configuration has been designed to provide maximum performance for engineering and scientific simulations. Each node provides a theoretical maximum of 410 GB/s for memory bandwidth limited codes and a theoretical maximum of 4.5 TFLOPs for compute limited codes.
We make SankhyaSutra Rudra available to users to meet their computation needs in HPC-as-Service mode.