0% found this document useful (0 votes)
50 views10 pages

Fcachememory

The document compares the performance of direct mapped, fully associative, and nested loop programs using different cache sizes and block sizes. It shows that fully associative memory has fewer misses than direct mapped memory. Increasing the block size from 1 to 4 reduces misses for both direct mapped and fully associative caches across different programs like nested loops and matrix multiplication.

Uploaded by

api-3738938
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as XLS, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
50 views10 pages

Fcachememory

The document compares the performance of direct mapped, fully associative, and nested loop programs using different cache sizes and block sizes. It shows that fully associative memory has fewer misses than direct mapped memory. Increasing the block size from 1 to 4 reduces misses for both direct mapped and fully associative caches across different programs like nested loops and matrix multiplication.

Uploaded by

api-3738938
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as XLS, PDF, TXT or read online on Scribd
You are on page 1/ 10

HITS & MISSES FOR DIRECT MAPPED USING PROGRAM

Cache % Avg.
Replacem Block
Program Sizes / Total Hits % Misses % Efficienc
ent size
Capacity8 234 196 84 38 16
y of hits
1 16 234 220 94 14 6 90.60
Nested 32 234 220 94 14 6
Direct loop,
Nil
mapped outer=12, 8 234 218 93 16 7
inner=4 4 16 234 230 98 4 2 96.58
32 234 230 98 4 2

Direct Mapped
Nested Loop (Blocksize 1)
110
100
90
80
70
60
Miss

50
40
30
20
10
0
Row 7 Row 8 Row 9
Hits

Column H Column J

Direct Mapped
Nested loop (BlockSize 4)
100
90
80
70
60
Miss

50
40
30
20
10
0
Row 11 Row 12 Row 13
Hits

Column H Column J
HITS & MISSES FOR DIRECT MAPPED MEMORY USING PROGRAM

Cache % Avg.
Replace Block
Program Sizes / Total Hits % Misses % Efficienc
ment size
Capacity y of hits
8 148 27 18 121 82
1 16 148 64 43 84 57 43.69
32 148 103 70 45 30
Direct Matrix
Nil
mapped Multiply 8 148 66 45 82 55
4 16 148 100 68 48 32 67.57
32 148 134 91 14 9

Direct Mapped
Matrix Multiply (Blocksize-1)
100
90
80
70
60
Miss

50
40
30
20
10
0
Row 6 Row 7 Row 8
Hits

Column H Column J

Direct Mapped
Matrix Multiply (Blocksize-4)
100
90
80
70
60
Miss

50
40
30
20
10
0
Row 10 Row 11 Row 12
Hits

Column H Column J
HITS & MISSES FOR FULLY ASSOCIATIVE MEMORY USING PROGRAM
% Avg.
Replace Block Sizes /
Program Total Hits % Misses % Efficienc
ment size Capacit
8 234 184 79 50 21
y of hits
1 16 234 220 94 14 6 88.89
Fully Nested 32 234 220 94 14 6
loop, LRU
Associative outer=12, 8 234 218 93 16 7
Memory inner=4 4 16 234 230 98 4 2 96.58
32 234 230 98 4 2

Fully Associative Memory By Using Nested Loop


(LRU) Blocksize-1
100
90
80
70
60
Miss

50
40
30
20
10
0
Row 6 Row 7 Row 8
Hits

Column H Column J

Fully Associative Memory By Using Nested Loop (LRU) Blocksize-4


100
90
80
70
60
Miss

50
40
30
20
10
0
Row 10 Row 11 Row 12
Hits

Column H Column J
HITS & MISSES FOR FULLY ASSOCIATIVE MEMORY USING PROGRAM
% Avg.
Replace Block Sizes /
Program Total Hits % Misses % Efficienc
ment size Capacit
y of hits
8 234 196 84 38 16
1 16 234 220 94 14 6 90.60
Fully Nested 32 234 220 94 14 6
loop
Associative outer=12,
Random
Memory 8 234 214 91 20 9
inner=4
4 16 234 230 98 4 2 96.01
32 234 230 98 4 2

Fully Associative Memory Nested loop


(Randam) Block Size = 1
100
90
80
70
60
Miss

50
40
30
20
10
0
Row 7 Row 8 Row 9
Hits

Column H Column J

Fully Associative Memory By Using Nested Loop (Randam) Blocksize-4


100
90
80
70
60
Miss

50
40
30
20
10
0
Row 11 Row 12 Row 13
Hits

Column H Column J
HITS & MISSES FOR FULLY ASSOCIATIVE MEMORY USING PROGRAM
Cache % Avg.
Replace Block
Program Sizes / Total Hits % Misses % Efficienc
ment size
Capacity y of hits
8 234 184 79 50 21
Fully Nested 1 16 234 220 94 14 6 88.89
loop 32 234 220 94 14 6
Associative outer=12
FIFO
Memory , inner=4 8 234 218 93 16 7
4 16 234 230 98 4 2 96.58
32 234 230 98 4 2

Fully Associative Memory By Using Nested Loop (FIFO) Blocksize-1


100
90
80
70
60
Miss

50
40
30
20
10
0
Row 7 Row 8 Row 9
Hits

Column H Column J

Fully Associative Memory By Using Nested Loop (FIFO) Blocksize-4


100
90
80
70
60
Hits

50
40
30
20
10
0
Row 11 Row 12 Row 13
Miss

Column H Column J
HITS & MISSES FOR FULLY ASSOCIATIVE MEMORY USING PROGRAM
% Avg.
Replace Block Sizes /
Program Total Hits % Misses % Efficienc
ment size Capacit
8 148 8 5 140 95
y of hits
1 16 148 48 32 100 68 36.71
32 148 107 72 41 28
Fully
Matrix LRU
Associative Multiply 8 148 82 55 66 45
Memory 4 16 148 90 61 58 39 67.57
32 148 128 86 20 14

Fully Associative Memory By Using Matrix Multiply (LRU) Blocksize 1


100
90
80
70
60
Miss

50
40
30
20
10
0
Row 6 Row 7 Row 8
Hits

Column H Column J

Fully Associative Memory By Using Matrix Multiply (LRU) Blocksize 4


100
90
80
70
60
Miss

50
40
30
20
10
0
Row 10 Row 11 Row 12
Hits

Column H Column J
HITS & MISSES FOR FULLY ASSOCIATIVE MEMORY USING PROGRAM
Cache % Avg.
Replace Block
Program Sizes / Total Hits % Misses % Efficienc
ment size
Capacity y of hits
8 148 21 14 127 86
1 16 148 55 37 93 63 39.41
Fully 32 148 99 67 49 33
Matrix
Associative Multiply Random
Memory 8 148 79 53 69 47
4 16 148 102 69 46 31 68.92
32 148 125 84 23 16

Fully Associative Memory By Using Matrix Multiply (Radom) Blocksize 1


100
90
80
70
60
Miss

50
40
30
20
10
0
Row 7 Row 8 Row 9
Hits

Column H Column J

Fully Associative Memory By Using Matrix Multiply (Random) Blocksize 4


100
90
80
70
60
Miss

50
40
30
20
10
0
Row 11 Row 12 Row 13
Hits

Column H Column J
HITS & MISSES FOR FULLY ASSOCIATIVE MEMORY USING PROGRAM
Cache % Avg.
Replace Block
Program Sizes / Total Hits % Misses % Efficienc
ment size
Capacity y of hits

8 148 8 5 140 95
Fully Nested 1 16 148 48 32 100 68 36.94
loop 32 148 108 73 40 27
Associative outer=12,
FIFO
Memory inner=4 8 148 82 55 66 45
4 16 148 90 61 58 39 67.57
32 148 128 86 20 14

Fully Associative Memory By Using Nested loop (FIFO Blocksize 1


100
90
80
70
60
Miss

50
40
30
20
10
0
Row 7 Row 8 Row 9
Hits

Column H Column J

Fully Associative Memory By Using Nested loop (FIFO) Blocksize 4


100
90
80
70
60
Miss

50
40
30
20
10
0
Row 11 Row 12 Row 13
Hits

Column H Column J
Cache
Replace Block
Program Sizes / Total Hits % Misses
ment size
Capacity8 234 196 84 38
1 16 234 220 94 14
Nested 32 234 220 94 14
Direct loop,
Nil
mapped outer=12, 8 234 218 93 16
inner=4 4 16 234 230 98 4
32 234 230 98 4

Cache
Replace Block
Program Sizes / Total Hits % Misses
ment size
Capacity8 148 27 18 121
1 16 148 64 43 84
32 148 103 70 45
Direct Matrix
Nil
mapped Multiply 8 148 66 45 82
4 16 148 100 68 48
32 148 134 91 14
% Avg.
% Efficienc
16
y of hits
6 90.60
6

7
2 96.58
2

% Avg.
% Efficienc
82
y of hits
57 43.69
30

55
32 67.57
9

You might also like