I am a newbie. I have NOT written a single line of CUDA code yet. I do understanding the benefit of running vector/matrix based calculation on a GPU. But my NEW apps will NOT calculation based. I need a lot of threads running in parallel. I do NOT want to upgrade my CPU/MOBO to have more cores. It is cheaper for me just to upgrade to a high end graphics card. Can I write C++ CUDA program so that I can run more threads in parallel so that the end result is it looks like I have CPU with A LOT OF CORES ?
Thanks for helping.