looking for simple cuda code


I am looking for a simple cuda code(c/c++) that explain what cuda is and what is good for.
I want that the code will show that the gpu works faster then the cpu.
for ex: the add of 2 vectors using cuda and without using cuda.

Can anyone tell where i can find such code?



The CUDA toolkit ships with various introductory examples. Often these include both a CPU based (“gold”) solution as well as the CUDA code, and in the end the output of CUDA is checked against the CPU code.

Some of these examples include timing measurements as well.

found it.
Can someone give me example of code that cuda is not efficient?
such as program for sequential calculating.