ADD file via upload

t4.cpp

@@ -0,0 +1,69 @@
+#include <stdio.h>
+#include <stdlib.h>
+#include <time.h>
+#include <arm_neon.h>
+
+#define SIZE 1024
+
+// NEON优化的稠密矩阵乘法函数
+void matmul_optimized(float** A, float** B, float** C, int n) {
+    for (int i = 0; i < n; i++) {
+        for (int j = 0; j < n; j++) {
+            float32x4_t sum = vdupq_n_f32(0.0);
+            for (int k = 0; k < n; k += 4) {
+                float32x4_t a = vld1q_f32(A[i] + k);
+                float32x4_t b = vld1q_f32(B[k] + j);
+                sum = vmlaq_f32(sum, a, b);
+            }
+            // 将结果累加到C[i][j]
+            float* sumPtr = (float*)&sum;
+            for (int l = 0; l < 4; l++) {
+                C[i][j] += sumPtr[l];
+            }
+        }
+    }
+}
+
+int main() {
+    // 分配内存并初始化矩阵
+    float** A = (float**)malloc(SIZE * sizeof(float*));
+    float** B = (float**)malloc(SIZE * sizeof(float*));
+    float** C = (float**)malloc(SIZE * sizeof(float*));
+
+    for (int i = 0; i < SIZE; i++) {
+        A[i] = (float*)malloc(SIZE * sizeof(float));
+        B[i] = (float*)malloc(SIZE * sizeof(float));
+        C[i] = (float*)malloc(SIZE * sizeof(float));
+    }
+
+    // 随机初始化矩阵A和B
+    srand((unsigned int)time(NULL));
+    for (int i = 0; i < SIZE; i++) {
+        for (int j = 0; j < SIZE; j++) {
+            A[i][j] = rand() % 100;
+            B[i][j] = rand() % 100;
+        }
+    }
+
+    // 计时开始
+    clock_t start = clock();
+    // 执行NEON优化的矩阵乘法
+    matmul_optimized(A, B, C, SIZE);
+    // 计时结束
+    clock_t end = clock();
+
+    // 输出运行时间
+    printf("NEON优化矩阵乘法耗时: %lf 毫秒\n", 1000.0 * (end - start) / CLOCKS_PER_SEC);
+
+    // 释放内存
+    for (int i = 0; i < SIZE; i++) {
+        free(A[i]);
+        free(B[i]);
+        free(C[i]);
+    }
+    free(A);
+    free(B);
+    free(C);
+
+    return 0;
+}