Plan links with clearer throughput, overhead, and timing estimates. Compare payload efficiency across frame sizes. Turn packet details into reliable network capacity decisions today.
Use application payload if you want goodput. Set L3/L4 overhead to zero if your payload already includes protocol headers.
These examples assume a 40-byte L3/L4 overhead, 14-byte MAC header, no VLAN tag, 4-byte FCS, 8-byte preamble, and 12-byte inter-frame gap.
| Link Speed | Application Payload | Overhead | Efficiency | Maximum Application Throughput |
|---|---|---|---|---|
| 100.00 Mbps | 256 B | 78 B | 76.65% | 76.65 Mbps |
| 1.00 Gbps | 512 B | 78 B | 86.78% | 867.80 Mbps |
| 1.00 Gbps | 1,500 B | 78 B | 95.06% | 950.57 Mbps |
| 10.00 Gbps | 9,000 B | 78 B | 99.14% | 9.91 Gbps |
Total on-wire frame bytes = application payload + L3/L4 overhead + MAC header + VLAN bytes + FCS + preamble/SFD + inter-frame gap.
Wire bits per packet = total on-wire frame bytes × 8.
Application efficiency = application payload ÷ total on-wire frame bytes.
Maximum application throughput = link speed × application efficiency.
Maximum packet rate = link speed ÷ wire bits per packet.
Requested wire bandwidth = entered packet rate × wire bits per packet.
Transfer time = total transfer bits ÷ delivered application throughput.
These formulas estimate line-rate behavior from packet structure, not only from nominal Ethernet port speed.
It estimates payload throughput, packet-rate ceilings, utilization, serialization time, overhead share, and transfer time. It focuses on on-wire Ethernet behavior rather than only headline port speed.
They consume line time on real Ethernet links. Ignoring them overstates usable throughput, especially when packets are small and packet rate is high.
Full duplex doubles aggregate send-plus-receive capacity, but each direction still has its own per-link rate. A 1 Gbps full-duplex link remains 1 Gbps each way, not 2 Gbps one way.
Fixed Ethernet and protocol overhead stay almost constant per packet. When payload shrinks, those fixed bytes consume a larger share of total line time.
The calculator caps delivered traffic at the physical maximum packet rate. That prevents impossible throughput values and shows the saturation point more realistically.
Yes, when 802.1Q tagging exists on the path you are modeling. Each VLAN tag adds 4 bytes, which slightly lowers payload efficiency.
Not exactly. It estimates payload delivery from frame structure and line rate. Retransmissions, congestion, shaping, loss, latency, and host performance can reduce real TCP goodput further.
Yes. Enter a larger application payload and keep overhead realistic. Jumbo frames usually improve efficiency because the fixed per-frame overhead is spread across more useful data.
Important Note: All the Calculators listed in this site are for educational purpose only and we do not guarentee the accuracy of results. Please do consult with other sources as well.