CS300-13_Parallel_Algorithms - Parallel Algorithms Sung...

Info iconThis preview shows pages 1–12. Sign up to view the full content.

View Full Document Right Arrow Icon
Parallel Algorithms Sung Yong Shin TC Lab CS Dept. KAIST
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Outline 1. Background 2. Parallel Computers 3. PRAM 4. Parallel Algorithms
Background image of page 2
1. Background Von Neumann Machines sequential machines executing one instruction at a time inherent limitation on computation speed imposed by the speed of electrical signals 1 ft / 1 nanosecond ( 10 -9 sec ) Parallelism or Concurrency: carrying out many operations simultaneously partition a complex problem in such a way that various parts of the work can be carried out independently and in parallel, and combine the results when all subcomputation are complete. need parallel computers to support this approach.
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Two approaches Hardware-oriented A parallel computer of a specific architecture is built. The parallel algorithms are developed to make use of hardware features to the best advantage. Problem-oriented Whether the parallel algorithms can truly enhance computation speed, or not. If so, how much ?
Background image of page 4
Issues (i) How to identify inherent parallelism practical limitations to build parallel computers abstracting ingredients from complex reality (ii) How to realize inherent parallelism suitable parallel algorithms parallel computer languages
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Inherent Parallelism probably not fully realized now but for the future !!! “ fascinating research topics” Nicholas Pippenger (1976) “ NC-class problems” ( N ick’s C lass ) “ ultra-fast on a parallel computer with a feasible amount of hardware” ( independent of the particular parallel model chosen ) Which problems can be solved substantially faster using many processors rather than one processor ? P NC P( n ) processors (log n ) m P = NC ?
Background image of page 6
Applications (needs ) Computer Graphics Databases Artificial Intelligence Computer vision / Image processing · · · · · · · ·
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
2. Parallel Computers SIMD ( S ingle I nstruction M ultiple D ata Stream ) MIMD ( M ultiple I nstruction M ultiple D ata Stream ) What does SISD stand for ?
Background image of page 8
Mutiply Branch Subtract Divide Add Function Unit Data Source x y Result x + y Program SISD
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Mutiply Branch Subtract Divide Add Function Unit Data Source x y Result x + y Program SIMD array processors vector processors (pipelining) Function Unit Function Unit Add Result s + q v w q s Result v + w Add Add
Background image of page 10
Branch Add Multiply Subtract Divide Data Source x y Process1 MIMD Function Unit Add Result s / q v w q s Result w + v Branch Multiply
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 12
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 02/04/2010 for the course COMPUTER S cs300 taught by Professor Unkown during the Spring '08 term at Korea Advanced Institute of Science and Technology.

Page1 / 31

CS300-13_Parallel_Algorithms - Parallel Algorithms Sung...

This preview shows document pages 1 - 12. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online