{{ result.summaryTitle }}
{{ result.primaryDisplay }}
{{ result.secondaryText }}
{{ badge.label }}
Enter the useful payload size before Ethernet headers, FCS, tags, padding, preamble, or inter-frame gap.
bytes
Use the Ethernet line rate for the path you want to model.
Pick untagged, 802.1Q, QinQ, or a custom tag count.
Set the number of 4-byte tags included between source MAC and EtherType.
tags
Adds 20 wire-time bytes per frame: 7-byte preamble, 1-byte start delimiter, and 12-byte inter-frame gap.
Leave at 0 to skip target-rate planning, or enter a desired goodput target.
Use 1500 for standard MTU analysis or 9000 for jumbo comparison.
bytes
MetricValueDetailCopy
{{ row.metric }} {{ row.value }} {{ row.detail }}
CheckStatusActionReasonCopy
{{ row.check }} {{ row.status }} {{ row.action }} {{ row.reason }}
PayloadWire bytesEfficiencyGoodputCopy
{{ row.payloadLabel }} {{ row.wireBytesLabel }} {{ row.efficiencyLabel }} {{ row.goodputLabel }} 
{{ formattedJson }}
Tags: Networking
Export to PDF Fullscreen
Customize
Advanced
:

Ethernet frame efficiency is the share of transmitted bytes that carry useful payload rather than framing overhead. A link can run at 1 Gbps, 10 Gbps, or faster while the payload goodput is lower because every frame also carries MAC addresses, a type or length field, optional VLAN tags, a frame check sequence, and physical-layer spacing between frames.

The difference is small for large payloads and large enough to matter for tiny packets. A standard 1500-byte payload spends most of its wire time on payload. A 46-byte minimum payload spends a much larger share on fixed overhead, so packet rate, forwarding load, and capacity planning can become the real constraint before the raw line rate looks full.

Payload bytes compared with repeated frame overhead Efficiency rises when fixed per-frame bytes are spread across a larger payload. MAC 14 B Tags 4 B each Payload the bytes counted as useful goodput FCS 4 B Preamble 8 B IFG 12 B payload efficiency = payload bytes / total wire bytes

Efficiency is not a judgment about whether Ethernet is fast or slow. It is a way to account for how much of the line is available to the payload under a chosen frame size and tagging pattern. The same line rate can carry different payload rates when the path uses untagged frames, 802.1Q tagging, stacked service tags, jumbo frames, or many small frames.

The most common mistake is to compare application throughput directly with nominal link speed. Ethernet framing, higher-layer headers, acknowledgments, queuing, retransmissions, storage, and application limits can all reduce a real transfer. Frame efficiency isolates only the Layer 2 byte accounting so that this part of the capacity model is explicit.

Technical Details:

An Ethernet MAC frame starts with a destination MAC address, source MAC address, and a type or length field. The fixed untagged MAC header is 14 bytes. Each 802.1Q-style tag adds 4 bytes, and the frame check sequence adds 4 bytes at the end of the MAC frame. Frames below the minimum MAC frame size are padded before transmission.

Wire-time accounting includes bytes that are not usually shown as part of the captured MAC frame. The preamble and start frame delimiter consume 8 bytes before the frame, and the inter-frame gap reserves 12 byte times after it. Counting those 20 repeated byte times is important when estimating line-rate capacity, especially for small frames.

Formula Core:

The efficiency model treats the entered payload as useful data and everything else as overhead for that frame. Tag count changes the header and minimum-frame size. The optional wire gap changes whether physical-layer spacing is included.

Btag = 4×T Bheader = 14+Btag Bpadding = max(0,64+Btag-(Bheader+P+4)) Bwire = Bheader+P+4+Bpadding+G efficiency = PBwire goodput = line rate×efficiency
Quantity Bytes How it affects efficiency
Payload P 1 to 9600 entered bytes The numerator of the efficiency ratio and the useful data counted as goodput.
Base MAC header 14 bytes Destination MAC, source MAC, and type or length field.
VLAN tags T 4 bytes per tag, 0 to 8 tags Raises the MAC header size and the minimum tagged frame size.
Frame check sequence 4 bytes Trailer used to detect frame damage.
Minimum-frame padding Added only when needed Keeps the MAC frame at the required minimum length for the selected tag count.
Preamble, SFD, and IFG G 20 bytes when included Adds 8 bytes before the frame and 12 byte times after it for wire-time planning.

The resulting payload goodput is the line rate multiplied by the efficiency ratio. Frame rate is the line rate divided by total wire bits per frame. Overhead bandwidth is the remaining line-rate capacity after payload goodput is removed. When a target payload rate is supplied, the required wire rate is the target divided by the current efficiency.

Output cue Rule Plain meaning
high efficiency Efficiency is 94% or higher. Payload dominates the frame byte count.
normal overhead Efficiency is at least 85% and below 94%. Overhead is visible but not unusually large for many Ethernet planning cases.
overhead heavy Efficiency is below 85%. Small payloads, padding, or tag and wire-gap assumptions are consuming a large share.
small frames Payload is below 512 bytes. Batching, larger payloads, or offload behavior may matter more than raw bandwidth.
target short Modeled payload goodput is below the entered target payload rate. The current frame size and line rate cannot carry the requested payload rate in this model.

Everyday Use & Decision Guide:

Use a standard 1500-byte payload for ordinary Ethernet MTU planning and 9000 bytes when you are comparing a jumbo-frame path. Keep Include preamble, SFD, and inter-frame gap on when the result will be used for line-utilization, packet-rate, or capacity planning. Turn it off only when you intentionally want MAC-frame-only byte accounting.

The frame profile is the fastest way to avoid tag mistakes. Untagged Ethernet wire time models a plain Ethernet frame, 802.1Q tagged wire time adds one 4-byte tag, and QinQ / double-tagged wire time adds two tags. Use Custom tags and wire overhead when a service-provider path, lab capture, or design review needs a specific tag count.

  • Set Link rate to the nominal line rate only when the link is dedicated to the traffic being modeled. Use a shaped or reserved rate when policy limits the path.
  • Use Target payload rate to check whether a desired goodput can fit at the current frame efficiency.
  • Raise Curve max payload to 9000 when you want the efficiency curve to show jumbo payload behavior beside standard MTU payloads.
  • Compare Wire Ledger with Frame Guidance before acting on the headline percentage. The ledger explains the byte count, while guidance flags small frames, tag overhead, wire-time scope, and target fit.
  • Use Efficiency Curve when the payload size is negotiable. The curve makes the fixed-overhead penalty easier to see than a single percentage.

A high efficiency result does not prove that an application will reach that payload rate. TCP/IP headers, tunneling, encryption, acknowledgments, loss, buffering, storage, and endpoint CPU still sit above this Layer 2 calculation. Treat the result as the Ethernet byte-accounting part of a larger throughput estimate.

Calculations run in the page from the numbers you enter. The tool does not generate test traffic or probe a live network path, so results should be checked against captures, interface counters, or a throughput test before being used as operational proof.

Step-by-Step Guide:

Work from the frame shape first, then use the result tabs to decide whether payload size, tagging, or line rate is the main planning concern.

  1. Enter Payload bytes. Use 1500 for a normal MTU payload, 9000 for a jumbo-frame comparison, or a small value such as 46 or 64 to inspect fixed-overhead behavior.
  2. Set Link rate and choose Kbps, Mbps, Gbps, or Tbps. The rate converts the efficiency ratio into payload goodput, overhead bandwidth, and frame rate.
  3. Choose Frame profile. If the preset matches the path, let it set the tag count and wire-gap assumption for you.
  4. Adjust VLAN tags only when the path differs from the preset. Each tag adds 4 bytes to every frame.
  5. Leave Include preamble, SFD, and inter-frame gap on for wire-time planning. Clear it only when you are comparing MAC frame sizes without physical-layer spacing.
  6. Open Advanced when you need a Target payload rate or a different Curve max payload. Values outside the supported payload, tag, and curve ranges are forced back into valid bounds.
  7. Read Wire Ledger for the exact byte accounting, then check Frame Guidance for small-frame, tag, wire-time, and target-fit flags.
  8. Use Curve Data, chart exports, CSV, or JSON only after the frame profile and wire-gap scope match the path you intend to document.

Interpreting Results:

Payload efficiency is the headline ratio. It answers how much of the selected frame's wire-time byte count carries the entered payload. Payload goodput translates that ratio into a rate at the selected link speed, and Overhead bandwidth shows the rest of the line rate consumed by headers, FCS, padding, and optional wire gap.

Total wire bytes is the best sanity check when two efficiency runs disagree. A larger payload usually improves efficiency because the fixed overhead stays mostly constant. Extra VLAN tags, preamble/SFD plus inter-frame gap, and padding for very small payloads all increase wire bytes without increasing payload bytes.

  • Frame rate at line rate rises when frames are small, which can matter for switch, firewall, virtual NIC, and host packet-processing limits.
  • MAC frame only, no preamble or IFG will show a higher percentage than full wire-time accounting. Use the same scope when comparing runs.
  • Wire rate for target appears only after a positive target payload rate is entered.
  • target fits means modeled payload capacity is at or above the entered target. It does not account for higher-layer overhead or real-world congestion.
  • overhead heavy is a cue to inspect payload size first, then tag count and wire-gap scope.

Use the curve as a shape check. The largest gains usually appear when moving away from very small payloads. Once payloads are already near standard or jumbo MTU sizes, extra payload bytes improve the percentage more slowly.

Worked Examples:

A 1500-byte payload on a 1 Gbps link with one 802.1Q tag and full wire-time accounting uses 18 bytes of MAC header and tag, 4 bytes of FCS, and 20 bytes of preamble/SFD plus inter-frame gap. The total is 1542 wire bytes, so payload efficiency is about 97.28%. The modeled payload goodput is about 972.76 Mbps, leaving about 27.24 Mbps as Ethernet overhead.

A 46-byte untagged minimum payload with wire-time accounting has a 14-byte MAC header, 4-byte FCS, no padding, and 20 wire-time bytes. The total is 84 wire bytes, so efficiency is about 54.76%. On a 1 Gbps link, that is about 547.62 Mbps of payload goodput before any higher-layer headers or packet-processing limits are considered.

A 9000-byte payload with two VLAN tags on a 10 Gbps link has 22 bytes of header and tags, 4 bytes of FCS, and 20 wire-time bytes. The total is 9046 wire bytes, so efficiency is about 99.49%. The same tag overhead that matters for tiny frames becomes almost invisible when the payload is this large.

For a target-rate check, a 900 Mbps payload target with the tagged 1500-byte example requires about 925.20 Mbps of wire bandwidth. If the link or policy limit is lower than that required wire rate, the target will be short even before application or transport overhead is added.

FAQ:

Why is the payload rate lower than the link rate?

The link rate includes every transmitted byte time. Payload goodput removes MAC header bytes, tag bytes, FCS, padding, and optional preamble/SFD plus inter-frame gap from the useful share.

Should I include preamble, SFD, and inter-frame gap?

Include them for line-utilization, capacity, packet-rate, and wire-time planning. Clear the checkbox only when you intentionally want MAC-frame-only accounting for frame-size comparison.

Why can I enter a payload below the Ethernet minimum?

Small payloads are useful for inspecting padding behavior. When the payload plus header and FCS is below the minimum MAC frame size for the selected tag count, the model adds padding before computing wire bytes.

What changes when I choose QinQ?

QinQ uses two tags in this model, so tag overhead becomes 8 bytes per frame. That increases the MAC header, total wire bytes, and required wire rate for a given payload target.

Can this measure real throughput on my network?

No. The calculation uses the values you enter and does not send packets. Use interface counters, packet captures, or throughput tests to measure a live path.

Why did my custom value snap back into range?

Payload bytes are bounded from 1 to 9600, VLAN tags from 0 to 8, and curve max payload from 64 to 9600. Values outside those bounds are normalized so the result remains valid.

Glossary:

Payload
The useful bytes carried by the Ethernet frame and counted in the efficiency numerator.
Goodput
The payload data rate after the selected Ethernet frame and wire-time overhead are removed.
MAC frame
The Ethernet frame from destination MAC through FCS, excluding preamble, start delimiter, and inter-frame gap.
FCS
Frame check sequence, the 4-byte trailer used for frame error detection.
SFD
Start frame delimiter, the 1-byte field that marks the end of the preamble and the start of the frame.
IFG
Inter-frame gap, the required idle byte time between Ethernet frames.
VLAN tag
A 4-byte tag inserted into the Ethernet header, commonly used for 802.1Q VLAN identification.
QinQ
A double-tagged Ethernet frame pattern often used by service-provider or segmented Layer 2 paths.

References: