Advanced Matrix Addition Mapping

Explore boundary guards, coalesced memory access, and dynamic grid scaling.

Configuration

MATRIX SIZE (N) 8

BLOCK SIZE 4

Row-Major (Coalesced)

Uncheck to see "Strided" Column-Major mapping (Bad for performance).

Guard Logic

Within Bounds

// i, j calculated as usual...
int i = blockIdx.x * blockDim.x + threadIdx.x;
int j = blockIdx.y * blockDim.y + threadIdx.y;
int index = i + j * N;

if (i < N && j < N) { // "Out of Bounds" Check
  c[index] = a[index] + b[index];
} /* Thread Idle */

IDs B(?,?), T(?,?)

i ?

j ?

Coalesced Index

i + j * 8

HOVER TO TRACE MEMORY PATHS