__global__ void global_mmul (int *A, int *B, int *C, int N)
{
	int i = N-1 - (blockIdx.y * blockDim.y + threadIdx.y);
	int j = blockIdx.x * blockDim.x + threadIdx.x;

	int i_part  = i % PARTITION_SIZE;
	int j_part = j % PARTITION_SIZE;

	int rowPerPart = N/PARTITION_SIZE;

	__shared__ int Apart[PARTITION_SIZE][PARTITION_SIZE];
	__shared__ int Bpart[PARTITION_SIZE][PARTITION_SIZE];

	AT(C, N, i, j) = 0;
	for (int n = 0; n < rowPerPart; n++)
	{
		Apart[i_part][j_part] = AT(A, N, i, n*PARTITION_SIZE + j_part);
		Bpart[i_part][j_part] = AT(B, N, n*PARTITION_SIZE + i_part, j);

		__syncthreads();

		for (int k=0; k<PARTITION_SIZE; k++)
			AT(C, N, i, j) +=  Apart[i_part][k]*Bpart[k][j_part];
	}


}