2D Convolution

题面

Write a program that performs a 2D convolution operation on the GPU with "valid" 边界条件：kernel 仅在与输入完全重叠的区域应用。输入与 kernel 以行优先一维数组表示，输出同样为行优先存储。输出尺寸：
output_rows = input_rows - kernel_rows + 1
output_cols = input_cols - kernel_cols + 1

Implementation Requirements

• Use only native features (external libraries are not permitted)
• The solve function signature must remain unchanged
• The final result must be stored in the array output

Examples

Example 1 与 Example 2 见站点页面示例。

Constraints

• 1 ≤ input_rows, input_cols ≤ 3072
• 1 ≤ kernel_rows, kernel_cols ≤ 31
• kernel_rows ≤ input_rows
• kernel_cols ≤ input_cols
• Performance is measured with input_cols = 3,072, input_rows = 3,072, kernel_cols = 15, kernel_rows = 15