Overview of Supercomputers by nit66238

VIEWS: 19 PAGES: 20

									Overview of
Supercomputers
   Presented by:
   Matt Lockwood
   &

   Mohammad Haghkar
Supercomputers
   The category of computers that includes the
    fastest and most powerful (most expensive)
    ones available at any given time.
   Designed to solve complex mathematical
    equations and computational problems very
    quickly.
What are They Used For

   Climate prediction & Weather forecasting
What are They Used For (cont.)


   Computational chemistry
   Crash analysis
   Cryptography
   Nuclear simulation
   Structural analysis
How Do They Differ From a
Personal Computer
   Cost
       range from $100,000s to $1,000,000s
   Environment
       most require environmentally controlled rooms
   Peripherals
       lack sound cards, graphic boards, keyboards, etc.
       accessed via workstation or PC
   Programming language
       FORTRAN
History

   Seymour Cray (1925-1996)
       Developed CDC 1604 – first fully transistorized
        supercomputer (1958)
       CDC 6600 (1965), 9 MFlops
       Founded Cray Research in 1972 (now Cray Inc.)
           CRAY-1 (1976), $8.8 million, 160 MFlops
           CRAY-2 (1985)
           CRAY-3 (1989)
    Early Timeline of Supercomputers
 Period      Supercomputer                      Peak speed                       Location

1943-1944   Colossus             5000 characters per second    Bletchley Park, England
1945-1950   Manchester Mark I    500 instructions per second   University of Manchester, England
                                 20 KIPS (CRT memory)          Massachusetts Institute of Technology,
1950-1955   MIT Whirlwind
                                 40 KIPS (Core)                Cambridge, MA
                                 40 KIPS
1956-1958   IBM 704
                                 12 kiloflops
                                 40 KIPS
1958-1959   IBM 709
                                 12 kiloflops
1959-1960   IBM 7090             210 kiloflops                 U.S. Air Force BMEWS (RADC), Rome, NY
1960-1961   LARC                 500 kiloflops (2 CPUs)        Lawrence Livermore Laboratory, California
                                 1.2 MIPS
1961-1964   IBM 7030 "Stretch"                                 Los Alamos National Laboratory, New Mexico
                                 ~600 kiloflops
                                 10 MIPS
1965-1969   CDC 6600                                           Lawrence Livermore Laboratory, California
                                 3 megaflops
1969-1975   CDC 7600             36 megaflops                  Lawrence Livermore Laboratory, California
                                 100 megaflops (vector),
1974-1975   CDC Star-100                                       Lawrence Livermore Laboratory, California
                                 ~2 megaflops (scalar)
                                 80 megaflops (vector),        Los Alamos National Laboratory, New Mexico
1975-1983   Cray-1
                                 72 megaflops (scalar)         (1976)
Where Are They Now

   www.top500.org
   List released twice a year
   Scores based on Linpack benchmark
   Solve dense system of linear equations
   Speed measured in floating point operations
    per second (FLOPS)
    Architectures - SMP
   Symmetric Shared-
    Memory
    Multiprocessing
    (SMP)
       Share memory
       Common OS
       Programs are divided
        into subtasks (threads)
        among all processors
        (multithreading)
    Architectures – MPP
   Massively Parallel Processing (MPP)
       Individual memory for each processor
       Individual OS’s
       Messaging interface for communication
       200+ processors can work on same application




        1. A large retailer wants to know how many camcorders the company sold
                                                                                          3. Each sub-query is assigned to a specific processor in the system. To
                in 1998, and sends that query to the MPP system.
                                                                                          allow this to happen, the database was previously partitioned. For
        2. The query goes out to one of the processors which acts as the                  example, a sales tracking database might be broken down by month, and
                coordinator, it breaks up the query for optimum performance. For
                example, it could break the query up by month; this “sub-query”             each processor holds data for one month’s worth of sales information.
                                                                                   4. The responses to the queries are returned to a processor to be coordinated—
               then goes to all the processors at the same time.
                                                                                            for example, each month is added up
                                                                                   5. Final answer is returned to the user.
Architectures – Clustering

   Grid computing
   Many servers connected together
   Relies heavily on network speed
   Easily upgraded with addition of more servers
Processor Types

   Vector processing
       Expensive
       NEC Earth Simulator
   Scalar processing
   Grid computing
       Based on off the shelf parts (ordinary CPUs)
BlueGene/L

   IBM
   MPP (massively parallel processing)
   #1 on top500 as of November 2004
   32,768 processors (700Mhz)
   70.72 Teraflops (trillions of FLOPS)
   Runs linux
   DNA, climate simulation, financial risk
   Cost more than $100 million
    BlueGene/L System Layout

   2 Processors
       Node communication
       Mathematical calculations
BlueGene/L Compute Card
BlueGene/L Node Board
BlueGene/L Cabinet
Some of the Others

   #2 - Columbia (NASA, USA) – 51.87 TFlops
   #3 - Earth Simulator (Japan) – 35.86 TFlops
   #4 - MareNostrum (Spain) – 20.53 TFlops
   #5 - Thunder (USA) – 19.94 TFlops
The Future
References
   http://www.top500.org/
   http://www.pcquest.com/content/Supercomputer/102051
    004.asp
   http://news.com.com/2100-1008_3-
    1000421.html?tag=fd_lede2_hed
   http://www.research.ibm.com/bluegene/index.html
   http://www.llnl.gov/asci/platforms/bluegene/papers/2hard
    ware_overview.pdf
   http://www.hpce.nec.com/451+M5f7cd421b8e.0.html
   http://www.cray.com/about_cray/history.html
   http://www.serc.iisc.ernet.in/~govind/243/L7-PA-Intro.pdf
   http://www.computerworld.com/hardwaretopic
    s/hardware/server/story/0,10801,43504,00.ht
    ml

								
To top