Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
SlideShare a Scribd company logo

Super COMPUTING Journal

•Download as PPTX, PDF•
•

This document discusses a supercomputer called HYPE-2 built by Santosh Pandey, Ram Sharan Chaulagain, and Prakash Gyawali under the supervision of Prof. Dr. Subarna Shakya. It provides an overview of multiprocessor and multicore systems and discusses how HYPE-2 uses a distributed memory architecture with dynamic scaling to achieve high performance computing capabilities for research applications like cryptography, data mining, and weather forecasting. Performance tests showed near-linear speedup as nodes were added, with the system able to handle complex computations through inter-process communication, though it is not as powerful as larger supercomputers.

1 of 36

1

Santosh Pandey Ram Sharan Chaulagain Prakash Gyawali - A supercomputer( HYPE -2 ) Supervisor Prof. Dr. Subarna Shakya

2

Super COMPUTING Journal

3

Super COMPUTING Journal

4

Super COMPUTING Journal

5

OUR OPTIONS: MULTIPROCESSOR SYSTEM MULTICOMPUTER SYSTEM

6

Super COMPUTING Journal

7

 Speedup in multiprocessing  Depends on parallelizable code S(P)=Speedup on P processors T(1)=Time to process in 1 processors T(P)= Time to process in processors f=Inherently sequential code p= Parallelizable code

8

 High performance computing for research  Achieving super computing at a cheaper rate than mainframes

9

 Muni Sakhya (1980’s)  16 nodes  First and the only one

10

Super COMPUTING Journal

11

 Middleware  Network Architecture  Multicore Computers

12

Super COMPUTING Journal

13

 SIMD (Single Instruction Multiple Data)  MIMD (Multiple Instruction Multiple Data)

14

 Every application don’t have same parallelism model  Specific Applications must be programmed  Extend Methods of our Architecture

15

Super COMPUTING Journal

16

Super COMPUTING Journal

17

 Dynamic Worker Addition and Reduction  Fault Tolerant  Scalable System

18

Super COMPUTING Journal

19

Star Topology

20

 Parallel Working

21

 Server Thread for each Worker at Server side  New Process for each Worker at Client side SERVER THREAD1 • Provide Chunk 1 To Client1 THREAD 2 • Provide Chunk 2 To Client2 THREAD N • Provide Chunk N To Client N Connect to server Take chunk to process Process Provide output to server Connect to server Take chunk to process Process Provide output to server Connect to server Take chunk to process Process Provide output to server

22

 Running thousands of flops operations  Integration for finding the value of Pi

23

Super COMPUTING Journal

24

Time (ms) 0 10000 20000 30000 40000 50000 1,4 1,3 2,1 3,1 3,2 3,3 4,3 Time (ms)

25

1(10000) 2(20000) 5(50000) 7(70000) 15(150000) Speedup 1 1.954 4.8 5.9 18 1 1.954 4.8 5.9 18 -5 0 5 10 15 20 25 Speedup No. of Nodes Speedup for 100 Million Iterations Fig : Exponential Speedup

26

1(10000) 2(20000) 5(50000) 7(70000)15(150000) Speedup 1 1.954 4.8 5.9 18 1 1.954 4.8 5.9 18 -5 0 5 10 15 20 25 Speedup No. of Nodes Speedup for 100 Million Iterations Theory vs. Practical Data

27

 No official data for comparing  Probably the fastest in Nepal

28

 Cryptography  Data Mining  Weather Forecasting  Research  Artificial Intelligence

29

Super COMPUTING Journal

30

Super COMPUTING Journal

31

Super COMPUTING Journal

32

Super COMPUTING Journal

33

 Not comparable with bigger super computer due to less nodes  Extension of Architecture library to define new application

34

 Supporting Complex Computations  Inter-process Communication for dependent tasks  Implementing GPU for Computation

35

Websites:  Don Berker. Robert G. Brown. Greg Lindahl. Forrest Hoffman. Putchong Uthayopas. Kragen Sitaker. Frequently Asked Questions [Online]. Available: http://www.beowulf.org/overview.faq.html  Technopedia. Computer Cluster [Online]. Available: http://www.technopedia.com/definition/6581/computer-cluster  Dr. Wu-chun. Feng. (2015). The Green500 list- November 2015 [Online]. Available: http://www.green500/list/green201511 Books:  Shiflet, Introduction to Computational Science: Modeling and Simulation for Sciences, Princeton University Press, 2014.  Kumar, Lenina, MATLAB: Easy Way to Learning, PHI Learning, 2016.  Etter, Introduction to MATLAB, Prentice Hall, 2015  Lemay Laura, Charles L. Perkins, Teach Yourself Java in 21 Days, Samsnet, 1996.

36

Super COMPUTING Journal

More Related Content

Super COMPUTING Journal

  • 1. Santosh Pandey Ram Sharan Chaulagain Prakash Gyawali - A supercomputer( HYPE -2 ) Supervisor Prof. Dr. Subarna Shakya
  • 5. OUR OPTIONS: MULTIPROCESSOR SYSTEM MULTICOMPUTER SYSTEM
  • 7.  Speedup in multiprocessing  Depends on parallelizable code S(P)=Speedup on P processors T(1)=Time to process in 1 processors T(P)= Time to process in processors f=Inherently sequential code p= Parallelizable code
  • 8.  High performance computing for research  Achieving super computing at a cheaper rate than mainframes
  • 9.  Muni Sakhya (1980’s)  16 nodes  First and the only one
  • 11.  Middleware  Network Architecture  Multicore Computers
  • 13.  SIMD (Single Instruction Multiple Data)  MIMD (Multiple Instruction Multiple Data)
  • 14.  Every application don’t have same parallelism model  Specific Applications must be programmed  Extend Methods of our Architecture
  • 17.  Dynamic Worker Addition and Reduction  Fault Tolerant  Scalable System
  • 21.  Server Thread for each Worker at Server side  New Process for each Worker at Client side SERVER THREAD1 • Provide Chunk 1 To Client1 THREAD 2 • Provide Chunk 2 To Client2 THREAD N • Provide Chunk N To Client N Connect to server Take chunk to process Process Provide output to server Connect to server Take chunk to process Process Provide output to server Connect to server Take chunk to process Process Provide output to server
  • 22.  Running thousands of flops operations  Integration for finding the value of Pi
  • 25. 1(10000) 2(20000) 5(50000) 7(70000) 15(150000) Speedup 1 1.954 4.8 5.9 18 1 1.954 4.8 5.9 18 -5 0 5 10 15 20 25 Speedup No. of Nodes Speedup for 100 Million Iterations Fig : Exponential Speedup
  • 26. 1(10000) 2(20000) 5(50000) 7(70000)15(150000) Speedup 1 1.954 4.8 5.9 18 1 1.954 4.8 5.9 18 -5 0 5 10 15 20 25 Speedup No. of Nodes Speedup for 100 Million Iterations Theory vs. Practical Data
  • 27.  No official data for comparing  Probably the fastest in Nepal
  • 28.  Cryptography  Data Mining  Weather Forecasting  Research  Artificial Intelligence
  • 33.  Not comparable with bigger super computer due to less nodes  Extension of Architecture library to define new application
  • 34.  Supporting Complex Computations  Inter-process Communication for dependent tasks  Implementing GPU for Computation
  • 35. Websites:  Don Berker. Robert G. Brown. Greg Lindahl. Forrest Hoffman. Putchong Uthayopas. Kragen Sitaker. Frequently Asked Questions [Online]. Available: http://www.beowulf.org/overview.faq.html  Technopedia. Computer Cluster [Online]. Available: http://www.technopedia.com/definition/6581/computer-cluster  Dr. Wu-chun. Feng. (2015). The Green500 list- November 2015 [Online]. Available: http://www.green500/list/green201511 Books:  Shiflet, Introduction to Computational Science: Modeling and Simulation for Sciences, Princeton University Press, 2014.  Kumar, Lenina, MATLAB: Easy Way to Learning, PHI Learning, 2016.  Etter, Introduction to MATLAB, Prentice Hall, 2015  Lemay Laura, Charles L. Perkins, Teach Yourself Java in 21 Days, Samsnet, 1996.