Supercomputing for MSMEs: The MSME Ministry has urged micro, small and medium enterprises to use the High-Performance Computing System (HPC) PARAM Utkarsh to reduce their turnaround time to market and increase innovation potential. Set up by Centre for Development of Advanced Computing (C-DAC) under the Ministry of Electronics and Information Technology (MeitY) as part of the National Supercomputing Mission (NSM) to carry out research and development activities in IT, electronics and associated areas, the HPC “offers Artificial Intelligence over machine learning and deep learning frameworks, compute and storage as a cloud service,” according to the details available on HPC portal.
HPC and storage services, big data cluster services, artificial intelligence over ML/DL, secured infrastructure, high stability and reliability, application enablement, and user support and training are the key features of PARAM Utkarsh.
Also read: Measuring ROI from technology implementations
The HPC pricing for MSMEs includes Rs 42 GPUcard/hour and Rs 0.96 CPUcore/hour while 100 GB storage is free per account per month and Rs 1 per GB thereafter. The billing period could be monthly, quarterly or annually. The system is available for commercial users and academic and R&D institutions.
“MSMEs are the backbone of the Indian economy, and many could use HPC to enhance their business. However, it can be challenging for MSMEs to adopt HPC. They may have little in-house expertise, limited access to hardware, or be unable to commit resources to a potentially risky endeavor,” the information available on HPC’s portal said.
Also read: Can warehouse automation technology unleash MSMEs’ real potential?
“This is where the National Supercomputing Mission (NSM) project comes in, by making it easier for MSMEs to try out their ideas for utilizing HPC to enhance their business, for example, to improve product quality, reduce time to delivery, or create innovative new services,” it noted.
PARAM Utkarsh is based on Intel Cascade Lake processor and NVIDIA Tesla V100 GPU with 100Gbps infiniband non-blocking interconnect. The system has over 50,000 compute cores (CPU & GPU), liquid cooling system for efficient PUE (power usage effectiveness) and offers peak computing power of 838 Teraflops.