Intel has as limit too… it is just high enough that almost nobody reaches it before they max out their CPU usage anyway.
As far as I can tell, Nvidia uses a much less efficient decoding method that is at least partially dependent on your GPU's processing power and available video memory. So a faster GPU with more video memory allows you to decode more cameras. I don't know what matters more... video memory or CUDA cores (overall GPU speed). From what I've seen so far, you would need a high-end gaming card like a GTX 1080 or t080ti in order for it to handle as much as Intel Quick Sync. This drastically increases the purchase price of the system, increases power consumption and heat generation, requires a bigger computer case and makes the computer significantly heavier. In fact you may need a bigger power supply to adequately feed such a huge GPU, and more system RAM because in
Blue Iris the Nvidia CUDA acceleration increases memory usage more than other decoding methods. For all of these reasons, Nvidia acceleration is
not a good choice for most installations.