Perhaps the most pressing challenge relating to DVR card performance is bottlenecking.
What is bottlenecking?
Say you have 100 garden hoses connected to each other, 99 of which can pump water at 3 gallons per minute, and one hose somewhere along the line, which can only pump water at 1 gallon per minute. In spite of all your 3 gallon-per-minute hoses, the entire system of hoses will only allow 1 gallon of water per minute to pass through.
The same is true in processing video data. “You’re only as fast as your slowest component.” Put a bunch of good high-performance components on a DVR encoder board and one component is not matched or as fast as the others, and you have a slow, underperforming system.
This brings us to a critical understanding with respect to performance. Many times manufacturers will provide a demonstration and show a limited number of channels of video performing at a certain level. Since the bottleneck will cause the solution to hit a wall at some point they limit the demonstration to stop before that critical point, where system degradation can be observed. So one needs to be careful to see what happens when you use the full capacity of the system.
Bottlenecking is but a single of multiple aspects of where the system may max out before its “alleged” capacity.