Why Per-Camera Bandwidth Is the Wrong Number to Size a Network Against
A single camera's bitrate — a few Mbps for a typical resolution and codec combination — seems trivially small relative to standard network link capacities. The relevant figure for network sizing is not any single camera's individual bitrate, but the AGGREGATE bandwidth of every camera streaming to the recorder simultaneously, since all cameras in a system typically transmit their recording streams continuously (or during active recording periods) to the same network video recorder or VMS server over the same shared network infrastructure.
Why Aggregate Bandwidth Scales Quickly With Camera Count
Even a modest per-camera bitrate multiplies into a substantial aggregate figure once camera count grows — a system of 32 cameras each streaming at a modest 4 Mbps produces 128 Mbps of aggregate bandwidth, and larger systems with higher resolution cameras or less aggressive compression can reach aggregate bandwidth figures well into the hundreds of Mbps or beyond, a very different sizing consideration than any individual camera's bitrate alone would suggest.
Why the Network Interface and Switch Uplink Both Need to Be Sized Above This Aggregate Figure
The network video recorder's own network interface card (NIC) has to be capable of receiving the full aggregate bandwidth from every camera simultaneously, and any switch uplink or network segment the camera traffic passes through before reaching the recorder needs equivalent or greater capacity — a bottleneck anywhere along this path, not just at the recorder itself, can cause dropped frames, recording gaps, or degraded live view performance, regardless of whether the recorder's own NIC is adequately sized.
Why Sustained Write Throughput Limits Practical Utilization Below Rated Link Speed
A nominally rated 1 Gigabit Ethernet (1GbE) link does not practically sustain a full 1,000 Mbps of continuous, real-world recording traffic — protocol overhead, network contention from other traffic sharing the link, and practical storage write performance limitations commonly mean a 1GbE link realistically sustains somewhere in the range of 700 to 800 Mbps of genuine continuous recording throughput before performance degrades. Sizing a network to the absolute theoretical maximum of its rated link speed, rather than this more realistic sustained throughput figure, risks a system that performs acceptably in isolated testing but degrades under real, continuous, multi-camera recording load.
Why Larger Systems Commonly Move to 10 Gigabit Ethernet
Once aggregate recording bandwidth approaches or exceeds the practical sustained throughput of a 1GbE link — commonly cited as roughly 500 Mbps as a reasonable planning threshold, providing margin below the approximately 700 to 800 Mbps practical ceiling — moving to 10 Gigabit Ethernet (10GbE) infrastructure for the recorder connection and relevant switch uplinks becomes the appropriate design response, providing substantially more headroom for the aggregate bandwidth of larger camera counts, higher resolutions, or less aggressive compression choices.
Why This Matters Even When Storage Capacity Alone Seems Adequate
It is possible to correctly size a storage array's total capacity in terabytes while still under-sizing the network infrastructure that has to deliver that data to the storage array in the first place — total storage capacity and network throughput capacity are two genuinely separate sizing questions that both need to be satisfied for a system to actually perform as designed. A storage array with abundant capacity is of little practical value if the network feeding it cannot sustain the aggregate write bandwidth the cameras are actually generating.
Why This Site's Calculator Reports Aggregate Bandwidth Explicitly
This site's Video Storage Sizing Tool explicitly calculates and displays aggregate bandwidth in both Mbps and Gbps, specifically so this network sizing question can be addressed directly alongside the storage capacity calculation, rather than treating network infrastructure as an afterthought considered only after storage capacity has already been finalized.