1 / 16

Parallel Computing I

Supercomputers Around The World. Nizamettin Doğan GÜNER 05.01.2011. CENG505. Parallel Computing I. Supercomputers. was introduced in the 1960s by Seymour Cray at Control Data Corporation

castropaul
Download Presentation

Parallel Computing I

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Supercomputers Around The World Nizamettin Doğan GÜNER 05.01.2011 CENG505 Parallel Computing I

  2. Supercomputers • was introduced in the 1960s by Seymour Cray at Control Data Corporation • Today, supercomputers are typically one of a kind custom designs produced by “traditional” companies as Cray, IBM and Hewlett Packard. • Supercomputers are used for highly calculation –intensive tasks such as problems involving quantum physics, weather forecasting, climate research, molecular modeling, physical simulations. Supercomputer is a computer that is the frontline of current processing capacity, particularly speed of calculation.

  3. Hardware and Software Design in Supercomputers Supercomputer Challenges, Technologies • Several technologies have been developed for supercomputers such as vector processing, liquid cooling, nonuniform memory access, parallelization. • A supercomputer generates large amount of heat and must be cooled. • A typical TOP 500 supercomputer consumes between 1 and 10 megawatt and converts all of it into heat. • The cost to power and cool the system is usually one of the factors that limit the scability of the system • *for example a high end system such as Tianhe-1A could consume several million dollars worth of electricity per year.)

  4. Operating Systems Supercomputers today most often usevariants of Linux as shown by the graph to the below.

  5. Programming and Software Tools • The base language of supercomputer code is, in general, Fortran or C, using special libraries to share data between nodes. • In the most common scenario, environments such as PVM and MPI for loosely connected clusters andOpenMP for tightly coordinated shared memory machines are used. • Significant effort is required to optimize a problem; the aim is to preventany of the CPUs from wasting time waiting on data from other nodes. • Software tools for distributed processing include standard APIs such as MPI and PVM, VTL, and open source-based software solutions such as Beowulf,WareWulf and openMossix. An easyprogramming language for supercomputers remains an open research topic in computer science. Several utilities that would once have cost several thousands of dollars are now completely free thanks to the open source community that often creates disruptive technology

  6. Modern Supercomputer Architecture • A computer cluster is a collection of computers that are highly interconnected via a high-speed network. • Each computer runs underan Operating System. • A multiprocessing computer is a computer, operating under a single OS and using more than one CPU, wherein the application-level software is indifferent to the number of processors. The processors share tasks using Symmetric multiprocessing (SMP) and Non-Uniform Memory Access (NUMA). MeasuringSupercomputerSpeed In general, the speed of a supercomputer is measured in "FLOPS“(FLoating Point Operations Per Second), “Petascale" supercomputers can process one quadrillion (1015) (1000 trillion) FLOPS.

  7. TOP 500 List • The TOP500 project ranks and details the 500 most powerfulknown computer systems in the world. • The project was started in 1993 and publishes an updated list ofthe supercomputers twice a year. Project website; www.top500.org • The TOP500 list is compiled by Hans Meuer of the University ofMannheim, Germany, Jack Dongarra of the University of Tennessee, Knoxville, Erich Strohmaier and Horst Simon of Lawrence Berkeley National Laboratory. • There is an application for the list of TOP 500 Supercomputers. This application allows you to visualise the list by the speed of each machine; the operating systems used; what it is used for; the country where it is based; the maker of the silicon chips used to build the machine and the manufacturer of the supercomputer. ‘http://www.bbc.co.uk/news/10187248’

  8. Visualizationof the TOP 500 List ‘Speed of each machine’

  9. Visualizationof the TOP 500 List ‘Operating systems used’

  10. Visualizationof the TOP 500 List ‘Usage Area’

  11. Visualizationof the TOP 500 List ‘Country where it is based’

  12. Visualizationof the TOP 500 List ‘The maker of the silicon chips used to build the machine’

  13. Visualizationof the TOP 500 List ‘Manufacturers of supercomputers’

  14. TOP 10 Supercomputers The following table gives the Top 10 positions of the 36th TOP500 List released on November 14, 2010

  15. Tianhe-1A is an upgraded supercomputer that is in China In October 2010, it overtook Jaguar to become the world's fastest supercomputer, with a peak computing rate of 2.507 petaFLOPSthat is, over 2½ quadrillion floating point operations per second. It has 186368cores. As announced on 13 November 2009, the Cray XT5 supercomputer known as Jaguar, knocked out the previous world number 1, IBM’s “Roadrunner”. The Jaguar’s theoretical processing peak is said to be 2.3 petaflopsper second. That’s 2.3 quadrillion calculations per second.It has 224162 cores. Columbia is a supercomputer built bySilicon Graphics for NASA. Its main purpose was to simulate the violent collision and merger of spiral galaxies that lead to the formation of elliptical galaxies. It entered the list in November 2004 at position 2, running at 51.87 teraflops, or 51.87 trillion floating point calculations per second.

  16. Thanks For Listening References • http://en.wikipedia.org/wiki/Supercomputer • http://www.top500.org/ • http://www.webopedia.com/TERM/S/supercomputer.html • http://tr.wikipedia.org/wiki/S%C3%BCper_bilgisayar • http://www.onlineticaret.net/site/page.asp?dsy_id=2624 • http://en.wikipedia.org/wiki/Top_500 • http://infosthetics.com/archives/2010/05/bbc_news_visualizing_the_top_500_supercomputer_report.html • http://www.bbc.co.uk/news/10187248

More Related