Updated: Mar 4, 2021
Supercomputers are computers with a high level of performance as compared to the general-purpose computers. They play an important role in the field of computational science, and are used for a wide range of computationally intensive tasks in various fields.
An ordinary computer does one thing at a time, so it does things in a distinct series of operations; that's called serial processing. A typical modern supercomputer works much more quickly by splitting problems into pieces and working on many pieces at once, which is called parallel processing. The most common supercomputer operating system used nowadays is Linux.
In 1964, Seymour Cray invented CDC-6600 which was 10 times faster than all computers of that time. Later in 1976, Cray delivered Cray-1 followed by Cray-2 in 1985. By 1993, Intel Paragon was the fastest supercomputer in the world. Presently, Fugaku in Japan is the fastest with a speed of 415 PFLOPS. Indian Institute of Tropical Meteorology in Pune hosts India’s fastest supercomputer Pratyush with a speed of 6.8 PFLOPS.
Supercomputers are used in a wide area of research and computation. The earliest Cray-1 supercomputer was used for weather forecast and aerodynamic research. As the advancement in supercomputer technology grew, they were started to be used in probabilistic analysis and 3D nuclear test simulations.
It was for this reason that the United States declined to offer a supercomputer to India fearing that it may be used for weapon development. Later India showed its technical prowess by indigenously developing PARAM 8000, which was 28 times more powerful than the one initially requested from the US. Created by C-DAC, PARAM 8000 was a success and was exported to Germany, UK and Russia. It was called the second most powerful supercomputer of its time.
These days supercomputers are also used for Molecular Dynamics simulations and for scientific research. More recently, supercomputers are being used to speed up the research on the coronavirus and vaccine development. They are being used in drug development for the coronavirus by employing them to look through databases of existing drug compounds. Various simulations are being run to find compounds that can stop the spread of the virus. Using multiple paralleled running CPU's, supercomputers are being run for hours non stop to model different processes. C-DAC in India collaborated with IITs, IISER and CSIR for performing drug repurposing simulations required towards the discovery of a new drug for COVID-19.
PARAM Padma, in 2002, became the first in India to achieve speed in teraFLOPS range. C-DAC plans to develop a supercomputer in the exaFLOPS range. The first supercomputer assembled indigenously, Param Shivay, was installed in IIT-BHU in 2019 followed by Param Shakti and Param Brahma at IIT-Kharagpur and IISER Pune, respectively.
The 860 TFLOP supercomputer in IIT Delhi is 166th fastest in the world. Based on NVIDIA’s Tesla platform, it is the largest High Performance Computing (HPC) system across all IITs. It is also the largest GPU-centric system in India and was recently upgraded in August last year which increased the raw CPU compute by 70% and GPU compute by about 40%.
The IIT Delhi supercomputer is widely used for research in molecular dynamics and virus dynamicity (to name a few). IIT Delhi also committed a total of 1 crore rupees worth of High Performance Computing (HPC) Resource for COVID-19 research. Presently, research is being carried out for finding a potential drug for the coronavirus and to find a “Plug & Play” solution that caters to surge in demand for ventilators in the face of medical emergencies such as COVID-19.
To quote experts, “Supercomputing has the potential to be the underlying layer to support solutions for many of the world’s most pressing contemporary challenges: global privacy and identity issues, stalemates in medical research, and sustainable supply-chain logistics, to name a few.” With more advanced technologies being developed to further increase the speed of supercomputers, a new age of research and development awaits us where the vaccine development would be faster and drug discovery would be easier. We will be able to better predict the weather, solve more complex algorithmic problems (like the traveling salesman problem), explore the edges of our universe, and create a more energy-smart power grid.