- Computer - Home
- Computer - Overview
- Computer - Advantages & Disadvantages
- Computer - Classification
- Computer - Applications
- Computer - History & Evolution
- Computer - System Characteristics
- Computing Origins
- Computer - Origins
- Computer - Generations
- Computer - Types
- Computer Components
- Computer - Components
- Computer - CPU
- Computer - CPU Components
- Computer - Arithmetic Logic Unit (ALU)
- Computer - Control Unit
- Computer - Input Devices
- Computer - Output Devices
- Computer - Memory Units
- Computer Hardware
- Computer - Hardware
- Computer - Motherboard
- Computer - GPU
- Computer - PSU
- Computer Software
- Computer - Software
- Computer - Software Types
- Computer Data Storage & Memory
- Computer - Data Storage
- Computer - Types of Data Storage
- Computer - Hard Disk Drives (HDD)
- Computer - Solid State Drives (SSD)
- Computer - Memory Cards
- Computer - Optical Storage
- Computer - NAND Flash Memory
- Computer - NVMe Memory
- Computer - CompactFlash Card
- Computer - Cloud Storage
- Computer Memory
- Computer - Memory
- Computer - Primary Memory
- Computer - Secondary Memory
- Computer - RAM
- Computer - ROM
- Computer - DRAM
- Computer - VRAM
- Computer - Cache Memory
- Computer Internet & Intranet
- Computer - Internet and Intranet
- Computer - Internet
- Computer - Extranet
- Computer - Websites
- Computer Office Tools
- Computer - Word Processors
- Computer - Spread Sheet
- Computer - Power Presentations
- Computer - E-mail Tools
- Computer Miscellaneous
- Computer - Ports
- Computer - Number System
- Computer - Number Conversion
- Computer - Data and Information
- Computer - Networking
- Computer - Operating System
- Computer - Keyboard Shortcut Keys
- Computer - Virus
- Computer - Antivirus
- Computer - How to Buy?
- Computer - Available Courses
- Calculator Vs. Computer
- Brain Vs. Computer
- Analog Vs. Digital Computer
- Supercomputer Vs. Mainframe Computer
- Mainframe Vs. Minicomputer
- Printer Vs. Scanner
- OCR Vs. MICR
- File Vs. Folder
- exFAT Vs. FAT32
- FAT32 Vs. NTFS
- FAT32 Vs. exFAT Vs. NTFS
- CPU Vs. GPU
- GUI Vs. CLI
- GUI Vs. CUI
- Cache Memory Vs. Virtual Memory
- Register Vs. Memory
- RAM Vs. CAM
- SRAM Vs. DRAM
- DDR2 Vs. DDR3
- PROM Vs. EPROM
- CD Vs. DVD
- Blu-Ray Vs. DVD
- MMU Vs. MPU
- Computer Terminologies
- Microsoft Windows
- Desktop
- Monitor
- Microphone
- Keyboard
- Keypad
- Mouse
- Printer
- Webcam
- Floppy Disk Drive
- Compact Disc
- Program
- Incognito Mode
- Electronic Mail (E-Mail)
- Server
- Cloud Hosting
- Minicomputers
- Supercomputer
- Bit
- Byte
- Kilobyte
- Megabyte
- Terabyte
- Yottabyte
- Zettabyte
- Exabyte
- Petabyte
- JEDEC
- Hub
- Block Storage
- Username
- URL
- Num Lock
- BIOS
- Bluetooth
What is Supercomputer?
A supercomputer is a special type of computer that is more powerful and capable of high-performance computing. It is specifically designed to compute complex and intensive tasks that regular computers cannot do efficiently.
Uses of Supercomputers
- Molecular mapping and research
- Weather forecasting
- Environmental research
- Oil and gas exploration
History of Supercomputer
The following is the history of supercomputer −
1960 − UNIVAC constructed the Livermore Atomic Research Computer (LARC) for the US Navy Research and Development Center. It is now recognized as one of the earliest supercomputers, relying on high-speed drum memory instead of the new disk drive technology.
1955 − Los Alamos National Laboratory requested a system 100 times faster than those existing at that time, leading to the creation of the IBM 7030 Stretch.
1961 − IBM completed the IBM 7030 Stretch, which utilized transistors, magnetic core memory, pipelined instructions, and random access disk drives. Despite not achieving the goal of a hundredfold speed increase, it was still purchased by the Los Alamos National Laboratory. The system was also sold in England and France and served as the basis for the IBM 7950 Harvest, a cryptanalysis supercomputer.
Early 1960s − The University of Manchester, under Tom Kilburn's leadership, built the Atlas supercomputer. It had 16,000 words in core memory, supplemented by a drum holding 96,000 words, and introduced time-sharing. The project was a collaboration between Ferranti and Manchester University, achieving processing speeds near one microsecond per instruction or one million instructions per second.
1964 − Seymour Cray completed the CDC 6600, marking the shift from germanium to silicon transistors. This allowed for faster operation, and refrigeration solved overheating issues. The CDC 6600 became the fastest computer in the world, surpassing others by about 10 times, and 100 units were sold at $8 million each, establishing the supercomputing market.
1972 − Seymour Cray left CDC to create his own company, Cray Research.
1976 − Cray Research released the Cray-1, running at 80 MHz, which became one of the most successful supercomputers in history.
1985 − The Cray-2 was released, featuring eight CPUs, liquid cooling, and Fluorinert as a coolant. It reached 1.9 gigaflops, becoming the first supercomputer to break the gigaflop barrier.
How do Supercomputers Work?
The following refers to the working model of supercomputers −
Architecture of Supercomputers
- Supercomputers function with architectures that consist of multiple central processor units (CPUs).
- These CPUs are arranged into clusters of computation nodes and memory storage.
- Supercomputers often have many nodes linked to solve problems by utilizing parallel processing.
Parallel Processing in Supercomputers
- The most powerful supercomputers are made up of multiple concurrent processors that conduct parallel processing.
- There are two types of parallel processing methodologies: symmetric multiprocessing and massively parallel processing.
- In some cases, supercomputers are dispersed, which means they take power from many PCs located in different areas, rather than placing all the CPUs in one location.
Measurement of Supercomputers' Performance
- Supercomputers are measured in floating point operations per second (FLOPS).
- Earlier systems were typically measured in instructions per second (IPS).
- The higher the FLOPS, the more efficient the supercomputer.
Organization of CPUs in Supercomputers
- Compared to regular computers, supercomputers have many CPUs.
- These CPUs are organized into compute nodes, each with a processor or group of processors — symmetric multiprocessing (SMP) — and a memory block.
- At scale, a supercomputer can contain a large number of nodes.
- These nodes work together to solve a specific issue using interconnect communications networks.
Power Consumption and Cooling in Supercomputers
- Due to the high power consumption of modern supercomputers, data centers require cooling systems.
- Suitable facilities are needed to accommodate all of the equipment.
Types of Supercomputers
Supercomputers are classified into the following types −
1. Tightly Connected Clusters
These are interconnected computers working together to solve a common problem. There are four methods to form clusters, which result in the following four cluster types: two-node clusters, multi-node clusters, director-based clusters, and massively parallel clusters.
2. Supercomputers with Vector Processors
This happens when the CPU processes an entire array of data items at once rather than one by one. This type of parallelism processes all array elements simultaneously. These processors are organized in arrays that handle many data items at the same time.
3. Special-purpose Computers
These are designed for a single function and cannot be used for other purposes. They are made to solve a specific problem and dedicate their resources to that task. The IBM Deep Blue chess-playing supercomputer is an example of such a system.
4. Commodity Supercomputers
These are composed of standard personal computers connected via high-bandwidth, high-speed Local Area Networks (LANs). They perform parallel computing to work on a single task.
5. Virtual Supercomputers
A virtual supercomputer operates in the cloud, providing an efficient computing platform by integrating multiple virtual machines on processors in a cloud data center.
Future of Supercomputers
The future of supercomputers looks very promising. As technology advances, supercomputers will continue to become more powerful and faster. They will be applied to solve complex challenges in science, engineering, data analytics, and artificial intelligence. Supercomputers will also play a role in powering self-driving cars, smart homes, and virtual reality experiences. Furthermore, they will be used for simulations and studies related to climate change effects.