The new supercomputer is expected to improve block-level weather forecasting, help meteorological scientists provide higher resolution ranges of the forecast, predict cyclones with greater accuracy and improve delivery times (the difference between a predicted phenomenon and what occurs actually) and provide information on ocean state forecasts, including seawater quality forecasts, PTI reported.
Currently we give forecasts with a resolution of 12 kilometers. The new supercomputer will improve it up to a resolution of six kilometers. Our goal is to achieve one kilometer resolution predictions, said the secretary of the Ministry of Earth Sciences, M Ravichandran. Making the announcement after a visit to the National Center for Medium-Range Weather Forecasts (NCMRWF) in Noida, Rijiju said the supercomputer will cost Rs 900 crore.
What are FLOPs in computer science?
FLOPs, or floating point operations per second, are a commonly used metric to measure the processing power and efficiency of computational performance, particularly in the fields of high performance computing (HPC) and artificial intelligence (AI) . Floating point operations are a certain type of mathematical calculation that uses real numbers with fractional parts.
How many FLOPs can a computer reach?
Modern computer systems, such as CPUs (Central Processing Units) and GPUs (Graphics Processing Units), are designed to perform multiple tasks simultaneously, using parallel processing techniques. Parallelism significantly increases the number of FLOPs a system can hit in a given amount of time. Over the years, hardware has become more efficient, exponentially increasing computing power.
For example, in 1961, the IBM 7030 Stretch, which cost a whopping $7.8 million at the time, performed a floating point multiplication every 2.4 microseconds, doing approximately 417,000 FLOPs. A PlayStation 5 today is listed to have peak performance of 10.28 TFLOPs, or 10.28 trillion FLOPs.
What is a petaFLOP?
Due to the immense computing power of today’s computers, the FLOP metric is often represented in terms of billions (gigs), trillions (teras), or even quadrillions (petas) of operations per second (GFLOP, TFLOP, PFLOP, respectively).
One petaflop is therefore equal to one thousand TFLOPs or 1015 FLOPs.
2008 was the first year that a supercomputer was able to break what was then called the petaFLOPS barrier, when the IBM Roadrunner shocked the world with an incredible peak performance of 1,105 petaFLOPS. Currently, the fastest computer in the world in terms of PFLOPs is the Hewlett Packard Enterprise Frontier, or OLCF-5 with the ability to hit a peak performance of 1685.65.
Are FLOPs the only metric to judge a computer’s performance?
No. FLOPs aren’t the only factor that determines the performance of a computer system. Memory bandwidth, latency, and other architectural characteristics also play a significant role. However, FLOPs provide a valuable baseline for comparing the computational capabilities of different systems, especially in tasks where floating point computations dominate.
Does India already use petaFLOP computers for weather forecasting?
NCMRWF hosts Mihir, a 2.8 petaflop supercomputer, while Indian Institute of Tropical Meteorology (IITM), Pune, hosts Pratyush, a 4.0 petaflop supercomputer, as per PTI. These were launched in 2018 and will be decommissioned once the new supercomputer is unveiled, a senior NCMRWF official told the news agency.
According to the agreement reached by the ministry, eight PFLOPs of computing power will be allocated to the NCMRWF, while the remaining 10 PFLOPs will go to the IITM. The Pune-based institute requires more power as it handles seasonal weather forecasts, while the NCMRWF handles medium-range forecasts for three to seven days in advance.
#India #Unveil #PetaFLOP #Supercomputers #Weather #Forecasting #FLOPs #Computing