matmultKernel.h

// matmultKernel.h
// For ECE-GY 9143 - High Performance Computing for Machine Learning
// Instructor: Parijat Dubey

// Kernels defined with this header must 
// multiply two matrices using CUDA: A x B = C

#ifndef __MMKERNEL__
#define __MMKERNEL__

// Defines for size of thread block and data computed by a thread block
#define BLOCK_SIZE 16
#ifndef FOOTPRINT_SIZE
#define FOOTPRINT_SIZE BLOCK_SIZE
#endif

// The type Matrix is really a MATRIX DESCRIPTOR. 
// Matrices are stored in row major order:
//       M[row,col] = *(M.elements + row * M.stride + col)
//
// A sub matrix is not copied but allocated in the full matrix.
//
// This requires the stride of the full matrix to properly get to the
// next row of the sub matrix (a block).
//
// Stride is the width in bytes from one element of the larger matrix 
// to the element in the same column but one row down.


typedef struct {
  int width;
  int height;
  int stride;
  float* elements;
} Matrix;

// Forward declaration of the kernel function that performs the work.
__global__ void MatMulKernel(const Matrix, const Matrix, Matrix);

#endif