Packet Rate Calculator – PPS, Bandwidth, and Packet Size
When sizing network hardware or diagnosing forwarding bottlenecks, raw bandwidth figures tell only part of the story. A router or firewall must process every individual packet, and its packets per second (PPS) forwarding capacity is often the real limiting factor — especially under small-frame or DDoS traffic. This calculator converts between bandwidth, PPS, and packet size so you can quickly answer capacity planning and traffic profiling questions.
The Core Formula
The fundamental relationship between bandwidth and PPS is:
PPS = Bandwidth (bps) ÷ (Packet Size + Overhead) × 8
Every packet transmitted on an Ethernet link carries a fixed per-frame overhead of 20 bytes: a 7-byte preamble, 1-byte Start Frame Delimiter (SFD), and 12-byte inter-frame gap (IFG). This overhead cannot be avoided and must be included in wire-rate calculations. Omitting it produces PPS figures that are approximately 1–30% too high depending on packet size.
Why Packet Size Matters So Much
PPS is inversely proportional to packet size — halving the packet size roughly doubles the required PPS. The table below illustrates this effect on a standard 1 GbE link with 20-byte Ethernet overhead:
| Packet Size | On-Wire Bytes | PPS at 1 Gbps |
|---|---|---|
| 64 B (min Ethernet) | 84 B | 1,488,095 |
| 128 B | 148 B | 844,594 |
| 512 B | 532 B | 235,048 |
| 1500 B (MTU) | 1520 B | 82,237 |
| 9000 B (jumbo) | 9020 B | 13,850 |
This is why security teams and hardware vendors describe DDoS attacks in PPS terms: an attacker flooding a link with minimum-size 64-byte frames generates ~18× more PPS than bulk file-transfer traffic at MTU, overwhelming the forwarding engine long before the bandwidth limit is reached.
Three Calculation Modes
The tool supports three directions of calculation to cover the most common engineering scenarios:
Bandwidth → PPS is the standard mode. Enter your link speed, average packet size, framing overhead, and optionally a link utilization percentage or number of bonded links. The result shows PPS in all scales (pps, Kpps, Mpps, Gpps) alongside the theoretical maximum wire-rate PPS for minimum-size frames, giving you the headroom margin.
PPS → Bandwidth is the reverse: given a forwarding target (for example, a hardware spec claiming 14.8 Mpps for a 10 GbE NIC), calculate how much bandwidth that represents at your expected packet size. This confirms whether the forwarding rate is achievable on a given link or if the bandwidth becomes the bottleneck first.
Reverse (Packet Size) derives the implied average packet size from two counter values collected in monitoring tools: observed bandwidth and observed PPS. The formula is:
Effective Packet Size = Bandwidth (bps) ÷ (PPS × 8)
A result significantly below the typical 512–1500 byte range indicates the link is dominated by small-frame traffic (control plane, DNS, or attack traffic), while a high value indicates bulk data transfers.
Multi-Link and Utilization Modeling
The Bandwidth → PPS mode supports two advanced inputs. The number of links field aggregates bonded interfaces (LACP bonds, multi-path ECMP) — enter 4 to model a 4×25 GbE bond as 100 Gbps total capacity. The link utilization slider models partial load; setting utilization to 60% calculates PPS at 60% of the nominal link speed, reflecting real-world traffic rather than theoretical wire rate.
Hardware Sizing Guidance
When selecting routers, switches, or firewalls, verify the device's PPS specification against the maximum wire-rate PPS for your link speed — not just the bandwidth rating. A device advertised as "1 Gbps capable" may only support 300–500 Kpps, which would be overloaded by small-frame traffic well before the bandwidth limit. Always add at least a 20–30% headroom over your calculated peak PPS when sizing production hardware.