Presentation is loading. Please wait.

Presentation is loading. Please wait.

Programming SA --- Systolic Array SIMD --- Single Instruction Multiple Data ISA --- Instruction Systolic Array MIMD --- Multiple Instruction Multiple Data.

Published by Modified over 9 years ago

Similar presentations

Presentation on theme: "Programming SA --- Systolic Array SIMD --- Single Instruction Multiple Data ISA --- Instruction Systolic Array MIMD --- Multiple Instruction Multiple Data."— Presentation transcript:

1 Programming SA --- Systolic Array SIMD --- Single Instruction Multiple Data ISA --- Instruction Systolic Array MIMD --- Multiple Instruction Multiple Data We will compare on examples the systolic merge architecture and the ISA merge architecture

2 parallel merge initial situation: 1.) sort columns (odd-even-transposition sort) 2.) sort rows (odd-even-transposition sort) sorted !!!! x1x2 x3 x4 x5 x6 x7 x17 x18 y1 y2 y3 y4 y5 y6 y7 y17 y18...

3 0-1 principle The 0-1 principle states that if all sequences of 0 and 1 are sorted properly than this is a correct sorter. The sorter must be based on moving data. initially 0s 1s after vertical sort 0s 1s after horizontal sort 0s 1s

4 MIMD-mesh (clocked) min maxTime: 2n

5 systolic merge 1 3 3 4 5 5 6 7 9 8 8 7 4 4 3 2

6 systolic merge 1 3 3 4 5 5 6 7 9 8 8 7 4 4 3 2

7 systolic merge 1 3 3 4 5 5 6 7 9 8 8 7 4 4 3 2

8 systolic merge 1 3 3 4 5 5 6 7 9 8 8 7 4 4 3 2

9 systolic merge 1 3 3 4 5 5 6 7 9 8 8 7 4 4 3 2 1 3 3 4 5 5 6 7 4 4 3 2 9 8 8 7

10 systolic merge 1 3 3 4 5 5 6 7 4 4 3 2 9 8 8 7

11 systolic merge 1 3 3 4 4 4 3 2 5 5 6 7 9 8 8 7

12 systolic merge 1 3 3 4 4 4 3 2 5 5 6 7 9 8 8 7

13 systolic merge 1 3 3 2 4 4 3 4 5 5 6 7 9 8 8 7

14 systolic merge 1 3 3 2 4 4 3 4 5 5 6 7 9 8 8 7

15 systolic merge 1 3 3 2 4 4 3 4 5 5 6 7 9 8 8 7

16 systolic merge 1 3 2 3 4 3 4 4 5 5 6 7 9 8 8 7

17 1 3 2 3 4 3 4 4 5 5 6 7 9 8 8 7 systolic merge

18 1 2 3 3 3 4 4 4 5 5 6 7 9 8 8 7 systolic merge

19 1 2 3 3 3 4 4 4 5 5 6 7 9 8 8 7 systolic merge

20 1 2 3 3 3 4 4 4 5 5 6 7 9 8 8 7 systolic merge

21 1 2 3 3 3 4 4 4 5 5 6 7 8 9 7 8 systolic merge

22 1 2 3 3 3 4 4 4 5 5 6 7 8 9 7 8 systolic merge

23 1 2 3 3 3 4 4 4 5 5 6 7 8 7 9 8 systolic merge

24 1 2 3 3 3 4 4 4 5 5 6 7 8 7 9 8

25 sorted !!! systolic merge 1 2 3 3 3 4 4 4 5 5 6 7 7 8 8 9

26 ISA merge 1 3 3 4 5 5 6 7 9 8 8 7 4 4 3 2 C:=min{C, CE} C:=max{C, CW} Array to be sorted

27 ISA merge 1 3 3 4 5 5 6 7 9 8 8 7 4 4 3 2

28 ISA merge 1 3 3 4 5 5 6 7 9 8 8 7 4 4 3 2

29 ISA merge 1 3 3 4 5 5 6 7 9 8 8 7 4 4 3 2

30 ISA merge 1 3 3 4 5 5 6 7 9 8 8 7 4 4 3 2

31 ISA merge 1 3 3 4 5 5 6 7 4 8 8 7 9 4 3 2

32 ISA merge 1 3 3 4 5 5 6 7 4 8 8 7 9 4 3 2

33 ISA merge 1 3 3 4 4 5 6 7 5 4 8 7 9 8 3 2

34 ISA merge 1 3 3 4 4 5 6 7 5 4 8 7 9 8 3 2

35 ISA merge 1 3 3 4 4 4 6 7 5 5 3 7 9 8 8 2

36 ISA merge 1 3 3 4 4 4 6 7 5 5 3 7 9 8 8 2

37 ISA merge 1 3 3 4 4 4 3 7 5 5 6 2 9 8 8 7

38 ISA merge 1 3 3 4 4 4 3 7 5 5 6 2 9 8 8 7

39 ISA merge 1 3 3 4 4 4 3 2 5 5 6 7 9 8 8 7

40 ISA merge 1 3 3 4 4 4 3 2 5 5 6 7 9 8 8 7

41 ISA merge 1 3 3 2 4 4 3 4 5 5 6 7 9 8 8 7

42 ISA merge 1 3 3 2 4 4 3 4 5 5 6 7 9 8 8 7

43 ISA merge 1 3 2 3 4 3 4 4 5 5 6 7 9 8 8 7

44 ISA merge 1 3 2 3 4 3 4 4 5 5 6 7 9 8 8 7

45 ISA merge 1 2 3 3 3 4 4 4 5 5 6 7 9 8 8 7

46 ISA merge 1 2 3 3 3 4 4 4 5 5 6 7 9 8 8 7

47 ISA merge 1 2 3 3 3 4 4 4 5 5 6 7 9 8 8 7

48 ISA merge 1 2 3 3 3 4 4 4 5 5 6 7 8 9 7 8

49 ISA merge 1 2 3 3 3 4 4 4 5 5 6 7 8 9 7 8

50 ISA merge 1 2 3 3 3 4 4 4 5 5 6 7 8 7 9 8

51 ISA merge 1 2 3 3 3 4 4 4 5 5 6 7 8 7 9 8

52 ISA merge 1 2 3 3 3 4 4 4 5 5 6 7 7 8 8 9

53 ISA merge 1 2 3 3 3 4 4 4 5 5 6 7 7 8 8 9

54 ISA merge 1 2 3 3 3 4 4 4 5 5 6 7 7 8 8 9

55 ISA merge 1 2 3 3 3 4 4 4 5 5 6 7 7 8 8 9

56 ISA merge 1 2 3 3 3 4 4 4 5 5 6 7 7 8 8 9

Download ppt "Programming SA --- Systolic Array SIMD --- Single Instruction Multiple Data ISA --- Instruction Systolic Array MIMD --- Multiple Instruction Multiple Data."

Similar presentations

Streaming SIMD Extension (SSE)

Streaming SIMD Extension (SSE)
EECC756 - Shaaban #1 lec # 1 Spring 2003 3-11-2003 Systolic Architectures Replace single processor with an array of regular processing elements Orchestrate.

EECC756 - Shaaban #1 lec # 1 Spring Systolic Architectures Replace single processor with an array of regular processing elements Orchestrate.
Batcher’s merging network Efficient Parallel Algorithms COMP308.

Batcher’s merging network Efficient Parallel Algorithms COMP308.
1 101418193716172325211 Merge and Count Merge and count step. n Given two sorted halves, count number of inversions where a i and a j are in different.

Merge and Count Merge and count step. n Given two sorted halves, count number of inversions where a i and a j are in different.
1 101418193716172325211 Merge and Count Merge and count step. n Given two sorted halves, count number of inversions where a i and a j are in different.

Merge and Count Merge and count step. n Given two sorted halves, count number of inversions where a i and a j are in different.
Heiko Schröder, 2003 Parallel Architectures 1 Various communication networks State of the art technology Important aspects of routing schemes Known results.

Heiko Schröder, 2003 Parallel Architectures 1 Various communication networks State of the art technology Important aspects of routing schemes Known results.
Application Specific Massive Parallelism Systolic and Instruction Systolic Computing Heiko Schröder, 2003 LOCOMAP.

Application Specific Massive Parallelism Systolic and Instruction Systolic Computing Heiko Schröder, 2003 LOCOMAP.
Mesh connected networks. Sorting algorithms Efficient Parallel Algorithms COMP308.

Mesh connected networks. Sorting algorithms Efficient Parallel Algorithms COMP308.
Презентація за розділом “Гумористичні твори”

Презентація за розділом “Гумористичні твори”
Центр атестації педагогічних працівників 2014

Центр атестації педагогічних працівників 2014
Галактики і квазари.

Галактики і квазари.
Характеристика ІНДІЇ.

Характеристика ІНДІЇ.
Процюк Н.В. вчитель початкових класів Боярської ЗОШ І – ІІІ ст №4

Процюк Н.В. вчитель початкових класів Боярської ЗОШ І – ІІІ ст №4
Selection Sort 345012 641532. 345012 6 41532 345012 64 1532.

Selection Sort
It is as easy as playing tic-tac-toe!. Make the vertical and horizontal lines of a tic- tac-toe game.

It is as easy as playing tic-tac-toe!. Make the vertical and horizontal lines of a tic- tac-toe game.
Flynn’s Taxonomy of Computer Architectures Source: Wikipedia Michael Flynn 1966 CMPS 5433 – Parallel Processing.

Flynn’s Taxonomy of Computer Architectures Source: Wikipedia Michael Flynn 1966 CMPS 5433 – Parallel Processing.
Determinants. Determinant - a square array of numbers or variables enclosed between parallel vertical bars. **To find a determinant you must have a SQUARE.

Determinants. Determinant - a square array of numbers or variables enclosed between parallel vertical bars. **To find a determinant you must have a SQUARE.
1 101418193716172325211 Counting Inversions Merge and count step. n Given two sorted halves, count number of inversions where a i and a j are in different.

Counting Inversions Merge and count step. n Given two sorted halves, count number of inversions where a i and a j are in different.
CA tracker for TPC online reconstruction CERN, April 10, 2008 S. Gorbunov 1 and I. Kisel 1,2 S. Gorbunov 1 and I. Kisel 1,2 ( for the ALICE Collaboration.

CA tracker for TPC online reconstruction CERN, April 10, 2008 S. Gorbunov 1 and I. Kisel 1,2 S. Gorbunov 1 and I. Kisel 1,2 ( for the ALICE Collaboration.
Flynn’s Taxonomy SISD: Although instruction execution may be pipelined, computers in this category can decode only a single instruction in unit time SIMD:

Flynn’s Taxonomy SISD: Although instruction execution may be pipelined, computers in this category can decode only a single instruction in unit time SIMD:

Similar presentations

Ads by Google