India’s first multi-petaflops Supercomputer

India’s first multi-petaflops Supercomputer 

Navneet Kumar Gupta 

India’s first multi-petaflops (PF) supercomputer inaugurated at Indian Institute of Tropical Meteorology (IITM) Pune on 08 January 2018 by Dr. Harsh Vardhan, Union Minister for Science & Technology, Earth Sciences, and Environment, Forest and Climate Change.

India’s first multi-petaflops supercomputer
As India takes the information superhighway towards realizing its digital dreams, it joins the ranks of a global consortium of IT-enabled nations that are lighting the path. One of the fields that have gained tremendous ground in the last few decades is Big Data and Big Data Analysis.

Therefore, India realized that it would have to not just build computers – but supercomputers to keep up. That fell to the technocrats at CDAC or the Central Department for Advanced Computing.

And so, in the late 1980s India’s finest scientists and computer engineers came together to build India’s first, completely indigenous supercomputer.

By 1991, they had built the PARAM 8000: A 1 Giga flop supercomputer. It was just the beginning. Over the next 3 decades, the PARAM series kept India at par with global supercomputing technology. Now that we have the technology, let’s understand: what can Big Data actually do for us? Big Data computation & analysis is the cornerstone of current-day weather forecasting. This is because today’s instruments to collect weather data are based not just on visual observation or measurement of things like temperature and pressure. Now India developed advance version of such supercompers and use them for Big Data analysis.
[post_ads]
Recently Union Minister of Earth Sciences Dr. Harsh Vardhan dedicates India’s first multi-petaflops supercomputer to the nation at IITM, Pune. In India The Ministry of Earth Sciences (MoES) in its endeavor to provide world class services (Weather, Climate, Ocean, Seismological) to the citizens of India, is constantly striving to upgrade its operational and research activities and other infrastructural facilities.

In order to provide useful services, MoES runs several computationally intensive applications on High performance computers (HPC). In last 10 years, the HPC infrastructure at MoES institutes has grown from 40 Tera flops in 2008 to 1 Peta flops in 2013/14. As a result of providing HPC infrastructure to MoES scientists, in addition to its constant research activity, MoES stands tall in the international weather and climate sciences community and constantly endeveours to provide the best services to the country and the Neighborhood. Some highlights of the services provided by MoES institutes to the country are:
  •  Best weather/climate prediction system in terms of accuracy/skill for monsoon climate. 
  •  Operational forecasts at a very high resolution of 3 km at regional scale and 12 km at global scale for weather forecasts at par with any other leading weather/climate forecast center in the world. 
  •  Fastest Tsunami alert/advisory to the stakeholders in Asia/Pacific region. 
  •  Advisories on Potential fishing Zones, operational ocean wave/weather watch forecasts and air quality forecasts, climate projections are provided regularly to the citizens of India.
MoES also plays leadership role in South Asian region by providing more accurate weather/climate forecasts to its neighboring countries.

Hence, constantly augmenting its High Performance Computing infrastructure at regular intervals is mandatory to keep MoES abreast with new technologies and to be at par with other leading weather/climate/ocean service centers worldwide.

The HPC facility inaugurated at IITM will be used for carrying out research on improving weather and climate forecasts and its applications. This MoES HPC facility will also be utilized by other MoES institutes [like Indian National Centre For Ocean Information Services (INCOIS), India Meteorological Department (IMD), National Institute of Ocean Technology (NIOT), National Centre for Antarctic and Ocean Research (NCAOR), The National Centre for Earth Science Studies (NCESS)] for research activities to improve their respective weather and climate services.

The second HPC facility at NCMRWF, Noida will be mainly used to cater daily operational forecasts of respective MoES institutes (INCOIS, IMD, IITM, NCMRWF).

This new HPC facility will enable MoES scientists to use more detailed components of the Earth System for making better weather and climate forecasts at very high resolution. The facility is expected to improve the following services:
  • Improved weather forecasts at block level over India which can predict extreme weather events. 
  • High resolution seasonal/extended range forecasts of active/break spells of Monsoon. 
  • Very high resolution coupled models for prediction of cyclones with more accuracy and lead time. 
  • Improved Ocean state forecasts including marine water quality forecasts at very high resolution. 
  • Tsunami forecasts with greater lead time. 
  • Air quality forecasts for different smart cities.
  • Climate projections at very high resolution.
This increase in supercomputing power will go a long way in delivering various societal applications committed by MoES. This will also give fillip to research activities not only in MoES but also in other academic institutions working on various problems related to Earth Sciences.
------------------
Navneet Kumar Gupta is a science communicator working as a Project Officer (Edusat) in Vigyan Prasar-National institute of Science communication under the Department. of Science & Technology. Govt. Of India. He has deep interest in popular science writing for general public through Print and electronic media. Besides his twelve books, he has written more than 200 popular science articles. He has edited/authored/co-authored more than 10 books. He have been awarded six National Awards including Rajbhasha Award, Ministry of Home Affairs, Government of India He has had a long stint as Associate Editor, VIPNET news - a popular science magazine. You may contact him at - ngupta@vigyanprasar.gov.in