惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

雷峰网
雷峰网
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
大猫的无限游戏
大猫的无限游戏
Google DeepMind News
Google DeepMind News
V
V2EX
T
The Blog of Author Tim Ferriss
H
Hackread – Cybersecurity News, Data Breaches, AI and More
Hugging Face - Blog
Hugging Face - Blog
Stack Overflow Blog
Stack Overflow Blog
I
InfoQ
博客园_首页
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
Last Week in AI
Last Week in AI
Recent Announcements
Recent Announcements
Vercel News
Vercel News
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
T
Tailwind CSS Blog
美团技术团队
Martin Fowler
Martin Fowler
宝玉的分享
宝玉的分享
Blog — PlanetScale
Blog — PlanetScale
GbyAI
GbyAI
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
J
Java Code Geeks

Louis C Deng's Blog

RoPE: Properties, Patterns, and Long-Context Behavior CS336 Assignment 1: Large Language Model Training and Inference CS231n Lecture Note: Generative Models CS231n Lecture Note: Self-Supervised Learning CS231n Lecture Note: Large Scale Distributed Training 自動微分 | DIY 實現自己的 PyTorch From RNNs to Transformers CS231n Lecture Note VII: Recurrent Neural Networks Uncovering Batch & Layer Normalization CS231n Lecture Note VI: CNN Architectures and Training CS231n Lecture Note V: Convolution Neural Networks Basics Demystifying Softmax Loss: A Step-by-Step Derivation for Linear Classifiers Backpropagation: A Vector Calculus Perspective CS231n Lecture Note IV: Neural Networks and Backpropagation CS231n Lecture Note III: Optimization CS231n Lecture Note II: Linear Classifiers CS231n Lecture Note I: Image Classification CSAPP Cache Lab I: Let's simulate a cache memory! CS188 Search Lecture Notes III CS188 Search Lecture Notes II How to Use TouchID for Sudo Commands on macOS CS188 Search Lecture Notes I RECAP2025: 留白 CSAPP Bomb Lab 解析 x64 暫存器速查表 CSAPP Data Lab 解析 矩陣的 Modified Gram Schmidt 方法 聊一聊位掩碼(Bit Mask) 整數溢位與未定義行為 快速排序 幾種劃分方法討論
CSAPP Cache Lab II: Optimizing Matrix Transposition
Louis C Deng · 2026-02-05 · via Louis C Deng's Blog

In this part of the Cache Lab, the mission is simple yet devious: optimize matrix transposition for three specific sizes: 32x32, 64x64, and 61x67. Our primary enemy? Cache misses.

Matrix Transposition

A standard transposition swaps rows and columns directly:

1
2
3
4
5
6
7
8
9
10
11
12
void trans(int M, int N, int A[N][M], int B[M][N])
{
int i, j, tmp;

for (i = 0; i < N; i++) {
for (j = 0; j < M; j++) {
tmp = A[i][j];
B[j][i] = tmp;
}
}

}

While correct, this approach is a cache-miss nightmare because it ignores how data is actually stored in memory.

Cache Overview

To optimize effectively, we first have to understand our hardware constraints. The lab specifies a directly mapped cache with the following parameters:

Parameter Value
Sets (S) 32
Block Size (B) 32 bytes
Associativity (E) 1 (Direct-mapped)
Integer Size 4 bytes
Capacity per line 8 integers

We will use Matrix Tiling and Loop Unrolling to optimize the codes.

32x32 Case

In this case, a row of the matrix needs 32/8 = 4 sets of cache to store. And cache conflicts occur every 32/4 = 8 rows. This makes 8x8 tiling the sweet spot.

By processing the matrix in 8×88 \times 8 blocks, we ensure that once a line of A is loaded, we use all 8 integers before it gets evicted. We also use loop unrolling with 8 local variables to minimize the overhead of accessing B.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
int i,j,k;
int tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7, tmp8;
for(i = 0; i<N; i+=8){
for(j = 0; j<M; j+=8){
for(k = i; k<N && k<i+8; k++) {

tmp1 = A[k][j];
tmp2 = A[k][j+1];
tmp3 = A[k][j+2];
tmp4 = A[k][j+3];
tmp5 = A[k][j+4];
tmp6 = A[k][j+5];
tmp7 = A[k][j+6];
tmp8 = A[k][j+7];


B[j][k] = tmp1;
B[j+1][k] = tmp2;
B[j+2][k] = tmp3;
B[j+3][k] = tmp4;
B[j+4][k] = tmp5;
B[j+5][k] = tmp6;
B[j+6][k] = tmp7;
B[j+7][k] = tmp8;
}
}
}

61x67 Case

Since 61 and 67 are not powers of two, the conflict misses don’t occur in a regular pattern like they do in the square matrices. This “irregularity” is actually a blessing. We can get away with simple tiling. A 16x16 block size typically yields enough performance to pass the miss-count threshold.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
int BLOCK_SIZE = 16;
int i,j,k,l,tmp;
int a,b;
for(i = 0; i<N; i+=BLOCK_SIZE){
for(j = 0; j<M; j+=BLOCK_SIZE){
a = i+BLOCK_SIZE;
b = j+BLOCK_SIZE;
for(k = i; k<N && k<a; k++) {
for(l = j; l<M && l<b; l++){
tmp = A[k][l];
B[l][k] = tmp;
}
}
}
}

64x64 Case

This is the hardest part. In a 64x64 matrix, a row needs 8 sets, but conflict misses occur every 32/8=432/8 = 4 rows. If we use 8x8 tiling, the bottom half of the block will evict the top half.

We can try a 4x4 matrix tiling first.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
int BLOCK_SIZE = 4;
int i,j,k,l,tmp;
int a,b;
for(i = 0; i<N; i+=BLOCK_SIZE){
for(j = 0; j<M; j+=BLOCK_SIZE){
a = i+BLOCK_SIZE;
b = j+BLOCK_SIZE;
for(k = i; k<N && k<a; k++) {
for(l = j; l<M && l<b; l++){
tmp = A[k][l];
B[l][k] = tmp;
}
}
}
}

But this isn’t enough to pass the miss-count threshold.

We try a 8x8 matrix tiling. We solve this by partitioning the 8×88 \times 8 block into four 4×44 \times 4 sub-blocks and using the upper-right corner of B as a “buffer” to store data temporarily.

Block A=(ATLATRABLABR)→TransposeBlock B=(ATLTABLTATRTABRT)\text{Block } A = \begin{pmatrix} A_{TL} & A_{TR} \\ A_{BL} & A_{BR} \end{pmatrix} \quad \xrightarrow{\text{Transpose}} \quad \text{Block } B = \begin{pmatrix} A_{TL}^T & A_{BL}^T \\ A_{TR}^T & A_{BR}^T \end{pmatrix}

Here are the steps:

  1. Transpose ATLA_{TL} into BTLB_{TL} while simultaneously moving ATRA_{TR} into BTRB_{TR} (as a temp storage).
  2. Move the stored ATRA_{TR} from BTRB_{TR} to its final position, while moving ABLA_{BL} into its spot.
  3. Transpose ABRA_{BR} into BBRB_{BR}.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
int i, j, k;
int tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7, tmp8;


for (i = 0; i < N; i += 8) {
for (j = 0; j < M; j += 8) {




for (k = 0; k < 4; k++) {

tmp1 = A[i + k][j]; tmp2 = A[i + k][j + 1];
tmp3 = A[i + k][j + 2]; tmp4 = A[i + k][j + 3];
tmp5 = A[i + k][j + 4]; tmp6 = A[i + k][j + 5];
tmp7 = A[i + k][j + 6]; tmp8 = A[i + k][j + 7];


B[j][i + k] = tmp1;
B[j + 1][i + k] = tmp2;
B[j + 2][i + k] = tmp3;
B[j + 3][i + k] = tmp4;



B[j][i + k + 4] = tmp5;
B[j + 1][i + k + 4] = tmp6;
B[j + 2][i + k + 4] = tmp7;
B[j + 3][i + k + 4] = tmp8;
}




for (k = 0; k < 4; k++) {

tmp1 = A[i + 4][j + k]; tmp2 = A[i + 5][j + k];
tmp3 = A[i + 6][j + k]; tmp4 = A[i + 7][j + k];


tmp5 = A[i + 4][j + k + 4]; tmp6 = A[i + 5][j + k + 4];
tmp7 = A[i + 6][j + k + 4]; tmp8 = A[i + 7][j + k + 4];


int t1 = B[j + k][i + 4];
int t2 = B[j + k][i + 5];
int t3 = B[j + k][i + 6];
int t4 = B[j + k][i + 7];


B[j + k][i + 4] = tmp1;
B[j + k][i + 5] = tmp2;
B[j + k][i + 6] = tmp3;
B[j + k][i + 7] = tmp4;


B[j + k + 4][i] = t1;
B[j + k + 4][i + 1] = t2;
B[j + k + 4][i + 2] = t3;
B[j + k + 4][i + 3] = t4;


B[j + k + 4][i + 4] = tmp5;
B[j + k + 4][i + 5] = tmp6;
B[j + k + 4][i + 6] = tmp7;
B[j + k + 4][i + 7] = tmp8;
}
}
}

Note: The key trick here is traversing B by columns where possible (so B stays right in the cache) and utilizing local registers (temporary variables) to bridge the gap between conflicting cache lines.

Conclusion

Optimizing matrix transposition is less about the math and more about mechanical sympathy—understanding the underlying hardware to write code that plays nice with the CPU’s cache.

The jump from the naive version to these optimized versions isn’t just a marginal gain; it’s often a 10x reduction in cache misses. It serves as a stark reminder that in systems programming, how you access your data is just as important as the algorithm itself.