The Top 10 supercomputers in the world June 2021
The Top10 supercomputers remained unchanged in the 57th edition of the TOP500. The sole new entry to the Top10 is the Perlmutter system at the DOE Lawrence Berkeley National Laboratory’s NERSC. The system is heterogeneous, GPU accelerated, and CPU exclusive nodes, based on the HPE Cray Shasta Architecture. Likewise, with a speed of 64.6 Pflop/s, Perlmutter ranked fifth on the new list.
Fugaku, a Japanese supercomputer, remained at the top of the list. Fugaku, a system co-developed by Riken and Fujitsu, with a 442 Pflop/s HPL benchmark score. This performance outperforms the No. 2 Summit by three times. However, Fujitsu’s proprietary ARM A64FX processor powers the computer. As a result of this feat, some have dubbed this machine the first “Exascale” supercomputer. However, with 2 Eflop/s on the new HPL-AI benchmark, Fugaku has already proven this unprecedented level of performance.
Aside from that, we saw a lot of Microsoft Azure and Amazon EC2 Cloud instances at the top of the list. Most importantly, Azure is used by Pioneer-EUS, which took the No. 24 places, and Pioneer-WUS2, which took the No. 27 spots. Moreover, you can see the Digital Jalandhar website for further information on this subject.
Here is a summary of the Supercomputers in the Top10:
- Fugaku remains the best system. It has a total of 7,630,848 cores, allowing it to attain an HPL benchmark score of 442 Pflop/s. This puts it three times ahead of the second-placed system on the list.
- Most importantly, Summit, an IBM-built system at the Oak Ridge National Laboratory (ORNL) in Tennessee, USA, continues to be the fastest system in the United States and the second-fastest system in the world, with a performance of 148.8 Pflop/s on the HPL benchmark.
- Sierra, a system at the Lawrence Livermore National Laboratory in California, USA, is ranked third. Its architecture is quite close to the architecture of the No. 2 system, Summit.
- With 93 Pflop/s, Sunway TaihuLight, a machine constructed by China’s National Research Center of Parallel Computer Engineering & Technology (NRCPC) and deployed at the National Supercomputing Center in Wuxi, Jiangsu province, is ranked No. 4. If you want to learn computers and have a hand in technology, you should seek help from the Computer Course Training in Jalandhar.
- Perlmutter is a newcomer to the TOP10 at No. 5. Perlmutter was able to attain a speed of 64.6 Pflop/s.
- Selene , now at No. 6, is an NVIDIA DGX A100 SuperPOD. The system is based on an AMD EPYC processor with NVIDIA. A100 for acceleration and a Mellanox HDR InfiniBand as a network and achieved 63.4 Pflop/s.
- Moreover, Tianhe-2A (Milky Way-2A), a system designed by China’s National University of Defense Technology (NUDT) and installed at the National Supercomputer Center in Guangzhou, is now ranked No. 7 with 61.4 Pflop/s.
- JUWELS Booster Module, Atos BullSequana system. Similar to the Selene System, the system employs an AMD EPYC processor with NVIDIA A100 acceleration and a Mellanox HDR InfiniBand network. With 44.1 Pflop/s, this system is the most powerful in Europe.
- HPC5 at No. 9 Dell PowerEdge system. Using NVIDIA Tesla V100 accelerators and a Mellanox HDR InfiniBand network, it reaches a performance of 35.5 Pflop/s.
- Dell C6420 system, is presently listed as the No. 10 system at the University of Texas’ Texas Advanced Computing Center. It used 448,448 Intel Xeon processors to reach 23.5 Pflop/s.
Top 20 countries Share:
The History of Supercomputers
- Does it make a difference how many CPUs or RAM you have? Is it necessary for a supercomputer to occupy a specific amount of space or to consume a specific amount of power?
- Moreover, the Control Data Corporation (CDC) 6600 was the first single-processor supercomputer.
- Moreover, The CDC 6600 was a small computer released in 1964. It was around four filing cabinets in size.
- It cost $8 million (about $60 million in today’s money) and ran at up to 40MHz, with a peak performance of 3 million floating-point operations per second (flops).
- Most importantly, The CDC 6600 had ten Peripheral Processors, each of which was responsible for I/O management and keeping the CPU’s queue full.
- Most importantly, Each of the ten parallel functional units in the CPU was dedicated to a certain task, such as floating-point add, floating-point divide, boolean logic, and so forth. The architecture employed was superscalar.
- The CPU featured 60-bit words and 60-bit registers, but it only had a limited instruction set because it only dealt with data that had been pre-processed by the Peripheral processor.
- Currently, It was because of this simplicity that the CDC 6600 CPU could be clocked so high.
- Seymour Cray designed the CDC 6600.
About Indian Supercomputers
1.Indigenous development programme
There is no denying the fact that in 1987 the Indian Government had requested to purchase a Cray X-MP supercomputer; this request was denied by the United States government as the machine could have a dual-use in weapons development. Most importantly, next to this issue, the Government of India decided to encourage an indigenous supercomputer development initiative the following year.
However, among the organisations commissioned to work on various projects were the Center for Advanced Computing Development (CDAC), the Center for Telematics Development (C-DOT), the National Aerospace Laboratories (NAL), The Bhabha Atomic Research Center (BARC) and the Advanced Numerical Research and Analysis Group (ANURAG).
2. C-DAC First Mission
Most importantly, the establishment of the centre Advanced Computing Development (C-DAC) was Between November 1987 and August 1988. Most importantly, In 1991, C-DAC debuted the PARAM 8000 supercomputer. Likewise, the applicants might choose a 120 hours computer course to overcome this test with flying colours. Moreover, this can certainly assist students to have a better understanding of advances.
3. C-DAC Second Mission
The main thing for C-DAC is the successful delivery of a gigaFLOPS parallel computer range of the PARAM 8000. Since 1992, the C-DAC has carried out its ‘second mission’ to deliver 100 GFLOPS line computers until 1997 and 1998.
Moreover, the aim was to make it possible for the computer to reach 1 teraFLOPS. 1993 experienced the release of the PARAM 9000 series of supercomputers with a top calculating capability of 5 GFLOPs. To deepen your grip on sophisticated technology and languages of the computers, you may connect to the best C++ language Coaching in Jalandhar.
4. C-DAC Third Mission
Firstly, the final task of C-DAC was to create a computer of the teraFLOPS range. In December most of the people experienced the supply of the PARAM Padma. C-DAC was included in the list of the world’s fastest supercomputers in June 2003. If you want to understand more about this subject also, see the Digital Jalandhar website for more additional information.
5. Development by other groups in the early 2000s
Most importantly, only ANURAG, BARC, C-DAC, and NAL continued to build their supercomputers in the early 2000s, according to reports. ANURAG continued to work on PACE, which was largely based on SPARC processors, at the same time. If you are also looking to learn the C language in the future then don’t search any further and plan to take help from the top C language Coaching in Jalandhar.
6. 12th Five Year Plan
In other words, during the 12th Five-Year Plan period (2012–2017), the Indian government has proposed allocating 2.5 billion dollars on supercomputing research. The Indian Institute of Science (IISc) in Bangalore would be in charge of the project. C-DAC will create it within the next five years upon proper permission.
7. National Supercomputing Mission
Firstly, the Ministry of Electronics and Information Technology launched a “National Supercomputing Mission” (NSM) in 2015, with the goal of installing 73 indigenous supercomputers by 2022. This is a $730 million, seven-year initiative (Rs. 4,500 crore). Previously, India built the computers; however, the NSM intends to manufacture the components in India. C-DAC and the Indian Institute of Science are implementing the NSM.
Moreover, the goal is to build a cluster of geographically dispersed high-performance computer facilities connected by a high-speed network, which will connect diverse academic and research organisations across India. Most importantly, the task entails setting up three petascale supercomputers, as well as capacity and capability machines.
In other words, the deployment of supercomputers with 60 per cent Indian components was the initial step. The second phase machines will come into existence from April 2021, the second phase. Within the NKN, the third and final phase aims to deploy completely indigenous supercomputers with a target performance of 45 petaFLOPS. AS of Nov-2020, there are three systems based in India on the top 500 list-