Open sarrvesh opened 8 years ago
Before doing RM Clean, it is probably useful to rotate the Q(\phi), U(\phi) and P(\phi) cubes such that \phi is the fastest varying axis. This can be done using the following code snippet:
index=0; for(x=0; x<qAxisLen1; x++) { for(y=0; y<qAxisLen2; y++) { for(z=0; z<qAxisLen3; z++) { newArray[index] = *(oldArray+ z*qAxisLen3 + y*qAxisLen2 + x); index += 1; } } }
Is it worth doing this rotation on the gpu than on the cpu?
Before doing RM Clean, it is probably useful to rotate the Q(\phi), U(\phi) and P(\phi) cubes such that \phi is the fastest varying axis. This can be done using the following code snippet:
index=0; for(x=0; x<qAxisLen1; x++) { for(y=0; y<qAxisLen2; y++) { for(z=0; z<qAxisLen3; z++) { newArray[index] = *(oldArray+ z*qAxisLen3 + y*qAxisLen2 + x); index += 1; } } }Is it worth doing this rotation on the gpu than on the cpu?