Which Are The Most Powerful Supercomputers In The Universe?

Source: Flickr

Definition Of A Supercomputer

A supercomputer is a name given to a group of computers with a lot of power. The fastest high-performance systems available at any given time are typically referred to by the term ‘Supercomputer’. These computers have primarily been used for scientific and engineering tasks that require extremely fast computations.

Testing mathematical models for complex physical phenomena or designs, such as climate and weather, the evolution of the cosmos, nuclear weapons and reactors, new chemical compounds (especially for pharmaceutical purposes), and cryptology, are some of the most common uses for supercomputers.

In the 1990s, more businesses started using supercomputers for market research and other business-related models as the cost of supercomputing fell. There are some distinguishing characteristics of supercomputers.

In contrast to conventional computers, they typically consist of more than one CPU (central processing unit), which houses circuits for correctly executing arithmetic and logic operations and interpreting program instructions. Due to the physical limitations of circuit technology, it is necessary to use multiple CPUs to achieve high computational rates.

Since the speed of light is the fundamental speed limit for signal transmission and circuit switching, electronic signals cannot travel faster than that. Due to the miniaturization of circuit components, dramatic length reductions in the length of wires connecting circuit boards, and advancements in cooling methods (for instance, processor and memory circuits in various supercomputer systems are immersed in a cryogenic fluid to achieve low temperatures at which they operate at their quickest), this limit has almost been reached.

To support CPUs' extremely high computational speed, it is necessary to quickly retrieve stored data and instructions. As a result, most supercomputers have very fast input/output speeds and a large storage capacity.

Still another distinctive trait of supercomputers is their utilization of vector math — i.e., They can work with lists of pairs of numbers rather than just pairs of numbers. A typical supercomputer, for instance, can, in roughly the same amount of time as it takes a standard computer to calculate the amount earned by just one worker, multiply a list of hourly wage rates for a group of factory workers by a list of hours worked by members of that group to produce a list of dollars earned by each worker.

Tasks related to national security, such as the design of nuclear weapons and cryptography, were the main initial uses of supercomputers. Supercomputers are also frequently used in the automotive, petroleum, and aerospace industries today.

In addition, supercomputers have found extensive use in engineering and scientific research, such as investigations into the structure of subatomic particles and the universe's origin and nature. In weather forecasting, supercomputers have become an indispensable tool:

Numerical models are now used to make predictions. Supercomputers became more widely used in online gaming as their prices fell. In particular, the Chinese supercomputers that were ranked fifth through tenth fastest in 2007 belonged to a company that owned the online rights to the electronic game World of Warcraft in China. At times, more than a million people played the game together at the same time.

How a Supercomputer’s Performance Is Measured? 

A supercomputer's performance is measured in exaflops, or one quintillion floating point operations per second, or at least 1018. The exa- prefix in exaflop refers to a quintillion, which is equal to one billion billion or 18 zeros. Similarly, a memory subsystem with a quintillion bytes of data is called an exabyte.

Floating point operations are represented by the abbreviation "flop" in exaflop. Exaflop/s is the rate at which a system executes a flop in seconds. Calculations in which all of the numbers are expressed with decimal points are referred to as floating points.

Here is a list of the three most powerful supercomputers  in the world right now

Source: Wikipedia

Frontier

In cooperation with its subsidiary Cray, the American multinational information technology company Hewlett Packard Enterprise built Frontier in 2022. Since it is the first exascale supercomputer in the world, it is capable of performing at least one quintillion (1018) calculations per second.

On Linpack benchmark tests, Frontier has a score of 1.1 EFLOPS(or exaflops)  and  8,730,112 cores. With a power efficiency rating of 52.23 gigaflops per watt, Frontier is also the most energy-efficient supercomputer in the world. The system's 74 computing cabinets weigh approximately 8,000 pounds (3.63 tons), and its total cost is $600 million.

Source: The Japan Times

Fugaku

Fujitsu, the famous Japanese IT giant, developed Fugaku in 2020 as a replacement for the company's 2011-built K computer. It was created to address the world's most pressing issues, particularly halting climate change.

It was Fugaku that occupied the spot of the fastest supercomputer in the world before Frontier took Fugaku's place in May 2022. Fugaku has 7,630,848 cores and a remarkable 442 PFLOPS (or petaflops) score, which indicates that it can perform 442 quadrillion calculations per second.

It has a power efficiency rating of just 14.78 gigaflops per watt and is powered by the Fujitsu A64FX 48-core 2.2GHz processor, making it quite inefficient by today's standards. The system in total has a cost of 700 Billion and each of the system cabinets weighs 1.6 tons and costs more than $1 billion.

Source: Wikipedia

Lumi

The fastest supercomputer in Europe, LUMI (Large Unified Modern Infrastructure), was built by HPE in 2022 and is located in Finland. The total number of cores in LUMI is 1,110,144 and it has a speed of 151.9 PFLOPS.

With a power efficiency rating of 51.63 gigaflops per watt, LUMI is the second most efficient supercomputer in the world and uses the same processor as Frontier.

Written By: Abderrahmane Loudiyi

Edited By: Nidhi Jha