These state-of-the-art machines will significantly improve forecasting capabilities at the block level, predict cyclones with greater accuracy and lead time, and provide detailed ocean state forecasts.
Floating-point operations involve mathematical calculations with real numbers that have fractional parts.
Using floating-point encoding, extremely long numbers can be handled relatively easily.
FLOPs are not the sole metric to evaluate a computer's performance. Factors like memory bandwidth, latency, and architectural features also contribute.
However, FLOPs provide a baseline for comparing computational capabilities, particularly in tasks dominated by floating-point calculations.
Unit of Computing Speed:
It is a unit of computing speed equal to one million million (1 trillion) (10^12) FLOPS.
It is a unit of computing speed equal to 1000 TFLOPS (10^15).
It is a unit of computing speed equal to one billion billion (10^18) FLOPS.
India's Current Usage of petaFLOPs:
The National Centre for Medium-Range Weather Forecasting (NCMRWF) houses 'Mihir,' a 2.8 petaFLOP supercomputer, while the Indian Institute of Tropical Meteorology (IITM) houses 'Pratyush,' a 4.0 petaFLOP supercomputer.
These existing supercomputers, launched in 2018, will be decommissioned once the new petaFLOP supercomputers are introduced.
As per the agreement, NCMRWF will receive eight PFLOPs of computing power, while the remaining ten PFLOPs will be allocated to IITM, catering to their specific weather forecasting requirements.
India’s first supercomputer called PARAM 8000 was launched in 1991.
The world’s fastest computer in terms of PFLOPs is the Hewlett Packard EnterpriseFrontier, or OLCF-5 with the capability to touch a peak performance of 1,685.65.
Airawat PSAI stands as India's largest and fastest AI supercomputing system, with a remarkable speed of 13,170 teraflops.