Software acceleration using GPU

Hello everyone,

I am new to GPU and Cuda programming. I have decent FPGA experience.
I have a custom software written in C++ (approx. 25k lines of code). I want to accelerate the software by 10X using GPU or FPGA. So far, I have not been very successful in implementing the software in FPGA (due to huge size of matrices and not adequate data precision) and now, I am thinking GPU might be better device.
I am looking for a reference design code (using cuda) that is implemented in a GPU for acceleration.
Can anyone please let me know where I can get the reference code?

Thank you,

Vivek.

The CUDA toolkit installer also installs sample codes which indicate how to accelerate various methods on the GPU.

Great! Thank you!