DOC PREVIEW
SJSU CS 147 - SuperComputers

This preview shows page 1-2-3-4-5 out of 16 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 16 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 16 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 16 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 16 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 16 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 16 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

Slide 1Overviewsōō'pər-kəm-pyōō'tərSupercomputerMeasuring PerformanceParallel Processing ArchitechuresCluster ArchitectureSDSU Supercomputers - DataStarSDSU Supercomputers - IntimidataVirginia Tech – System XUsesMultiprocessor ArchitectureSlide 13AMD Phenom – Quad Core…..Thank youFurther Info.- Rohan DhamnaskarOverviewWhat is a SupercomputerSome ConceptsCouple of examples.sōō'pər-kəm-pyōō'tərSupercomputerDictionary definition: a very fast, powerful mainframe computer, used in advanced military and scientific applications. For our sake: Computer SYSTEM designed to operate at super high speeds which can perform complex calculations that would require greater than a year for a normal computer.Measuring PerformanceFLOPS: FLoating point Operations Per SecondUsually prefixed by an SI unit of magnitude○ Megaflops, Gigaflops, Teraflops, Petaflops○Usual 3GHz processor ~ 6 GFLOPS.LINPACK: A Benchmark software library written in FORTRAN to compute a n x n systems of Linear Algebra. (MATH 129A!) The result is reported in MFLOPS.Parallel Processing ArchitechuresCluster ArchitectureEach Computer (Node) is a complete ComputerConnected to each other via a high speed busJob broken up into blocks of data and executed independently.Cheap and easy to produce.SDSU Supercomputers - DataStarIBM Power4+ basedPerformance: 15.6 teraflops (Overall)~ 6.2 teraflops each (4 separate cluster sets)Processors: 2528 IBM p-Series (1.5 ~ 1.7 GHz)Memory: 7.3 TB Storage: 180 TBSDSU Supercomputers - IntimidataIBM BG/L eServer (Blue Gene System)Performance: 17.2 teraflops (Overall)13.83 teraflops (LINPACK)Processors: 6144 PowerPC (700 MHz / 2 processors per Node)Memory: 1.5 TB Storage: 25 TBVirginia Tech – System X1110 Xserve G5 serversPerformance: 12.25 teraflops (Overall)10 teraflops (LINPACK)Processors: 2200 PowerPC (2.3 GHz / 2 processors per Node)Memory: 4.3 TB Storage: 16 TBCHEAP!!!UsesSupercomputers are mainly used to simulate:DNA / Protein Structures (Biology / Bioinformatics)Molecular Dynamics (Chemistry)Fluid Dynamics and Turbulence (Physics)Earthquake and Seismic activity (Geology)Fourier Transforms and Algebraic Systems (Mathematics)Global warming patterns (Environmental Sciences)Virtual CAD/CAM Design (Engineering)Solar atmosphere and Supernova Dynamics (Astronomy)NO COMPUTER SCIENCE ???Multiprocessor ArchitectureI/O PortDeviceDeviceControllerCPUMemoryCPUCPUBUSMultiprocessor Architecture2 to 8 processors (cores)Communication via Memory and System busLarge memory. (Memory Pool)Still Cheaper but lacks performance.Uses: Workstations, Application Servers, Gaming!!!AMD Phenom – Quad CoreProcessor Speed: 2.3 GHz (2 MB L2 + 2 MB L3 on chip)Processor – Processor Bandwidth: 27.2 GB/sFront Side Bus: Upto1066MHzUses : Gaming / Server…..Thank youFurther Info.VirginiaTech Advanced Reasearch Computing Center. < http://www.arc.vt.edu/arc/ >San Diego Supercomputing Center. < http://www.sdsc.edu/ >AMD Multicore Technology. < http://multicore.amd.com/ >LINPACK. < http://www.netlib.org/linpack/ >Lecture Notes: Parallel Computing. Dr.


View Full Document

SJSU CS 147 - SuperComputers

Documents in this Course
Cache

Cache

24 pages

Memory

Memory

54 pages

Memory

Memory

70 pages

Lecture 1

Lecture 1

53 pages

Cisc

Cisc

18 pages

Quiz 1

Quiz 1

4 pages

LECTURE 2

LECTURE 2

66 pages

RISC

RISC

40 pages

LECTURE 2

LECTURE 2

66 pages

Lecture 2

Lecture 2

67 pages

Lecture1

Lecture1

53 pages

Chapter 5

Chapter 5

14 pages

Memory

Memory

27 pages

Counters

Counters

62 pages

Load more
Download SuperComputers
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view SuperComputers and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view SuperComputers 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?