code example
Brook+ ATI GPU stream computing



// kernel
kernel void sub1(float x[], float y[],
  out float xout<>, out float eout<>)
{
  // do computation
  float i = 1 + indexof(xout).x;
  xout = x[i] + (y[i+1] + y[i-1])*.5f;
  eout = y[i] * y[i];
}

reduce void sub2(float y<>, reduce float e<>)
{
  e += y;
}

int main(int argc, char *argv[]) {
  int n = ...;
  float *x, *y;
  x = new float[n+1];
  y = new float[n+1];

  ... // fill x, y

  #define BLOCK (8190)
  float xStream<BLOCK>;
  float xoutStream<BLOCK>;
  float yStream<BLOCK+2>;
  float einStream<BLOCK+2>;
  float eStream<1>;
  float e = 0;
  float ee = 0;
  for (int i=1; i<n; i+=BLOCK) {
   streamRead(xStream, x+i);
   streamRead(yStream, y+i-1);
   sub1(xStream, yStream, xoutStream, einStream);
   streamWrite(xoutStream, x+i);
   sub2(einStream, eStream);
   float e_local;
   streamWrite(eStream, &e_local);
   ee += e_local;
  }
  e += ee;

  ... // output x, e

  delete[] x, y;
  return 0;
}

[start] [references] [download] [install]