Example for data broadcast

Accelerated Computing CUDA CUDA Programming and Performance
1

Hi,

 Has anybody used data broadcast for shared memory to avoid bank conflict? Its same shared memory access for multiple threads. 
                     
   If somebody has an example for the same . Please let me know.

Regards,
Ashok

2 
__global__ void mykernel()

{

	__shared__ int i;

	int j;

	i = 5; /* Brodcast write */

	j = i; /* Broadcast Read */

}