CUDA parsing of text files

I have a CUDA program already written in C++ for running on an NVIDIA processor. The program currently works fine, but I need to make some alterations to it. The former programmer no longer is available for the project as he has a full time job. The current program is written with Visual Studio for Windows. The alterations I wish to include involve some modest, additional, parsing of the data prior to the final output. Currently the program works on a single GPU, but it would be optionally beneficial if the program could use multiple GPUs if available. Should the interaction and project goals be met, there would be additional code I would want written. I am happy to spend more time working with the programmer to explain the requirements. Knowledge of bioinformatics/biology/DNA sequences would possibly help, but is not a necessity.