212
edits
Changes
→Original Problems when converting a program to use the GPU
The next problem encountered is like a vector we cannot transfer a 2D array to a *cudaMalloc* or *cudaMemCPy* function call. So the 2D array needs to be converted into a single array. And that brings another problem regarding the original algorithm which is shown below.
'''for (int i=0; i <= nx-1; i++)
{
if (i*dx >= 0.5 && i*dx <= 1)
''' for (int k = 0; k <= nx - 1; k++)
if (k*dx >= 0.5 && k*dx <= 1)
{
u[k * nt + 0] = 1;
}
for (int it = 1; it <= nx - 1; it++) { for (int k = 0; k <= nx - 1; k++) { un[k * nx + it - 1] = u[k * nx + it - 1]; } for (int m = 1; m <= nx - 1; m++) { u[0 * nx + it] = un[1 * nx + it - 1]; u[m * nx + it] = un[m * nx + it - 1] - c*dt / dx*(un[m * nx + it - 1] - un[(m - 1) * nx + it - 1]); }
}'''