Network Latency Calculator
Estimate round-trip time from propagation delay, serialization delay, switch/router hops, queueing buffer, packet size, link speed, measured RTT, and jitter budget.
🖧Network path presets
⚙Path inputs
Use cable route length, ISP path distance, or approximate server distance.
Include Ethernet overhead when comparing real serialization delay.
Enter 0 to use the calculated RTT only.
📶Network media and device grid
📊Propagation and media reference
| Media or path | Velocity factor | One-way delay | Practical note |
|---|---|---|---|
| Single-mode fiber | 0.67c | About 4.98 us/km | Backbone, metro, long LAN trunks |
| Copper Ethernet | 0.65c | About 5.13 us/km | Short runs, serialization matters more |
| WiFi 6 / 6E | 0.99c plus airtime | Low distance delay | Contention and retries dominate jitter |
| DOCSIS cable | Hybrid fiber coax | Plant plus scheduler | Upstream queues can dominate RTT |
| VPN tunnel | Underlying path | Crypto plus route | Adds processing and longer routing |
| GEO satellite | Radio path | About 250 ms one-way | Distance dominates everything else |
🔀Hop and device delay reference
| Device / segment | Typical added delay | Jitter risk | When to model it |
|---|---|---|---|
| Store-and-forward switch | 3-30 microseconds | Very low | Home lab switches and ToR paths |
| Router or firewall | 0.2-2 ms | Low to medium | NAT, inter-VLAN, IDS, shaping |
| WiFi AP airtime | 1-10 ms | Medium | Shared channels or weak clients |
| VPN endpoint | 1-8 ms | Medium | Small CPUs, encryption, overlay routes |
| ISP edge queue | 2-100 ms | High | Uploads near line rate or bufferbloat |
| Cloud load balancer | 0.5-5 ms | Low | Public services and region ingress |
⏱Packet serialization table
| Link speed | 64 byte frame | 1518 byte frame | 9000 byte jumbo |
|---|---|---|---|
| 100 Mbps | 0.005 ms | 0.121 ms | 0.720 ms |
| 1 Gbps | 0.001 ms | 0.012 ms | 0.072 ms |
| 2.5 Gbps | 0.0002 ms | 0.005 ms | 0.029 ms |
| 10 Gbps | 0.0001 ms | 0.001 ms | 0.007 ms |
| 100 Gbps | 0.00001 ms | 0.0001 ms | 0.0007 ms |
🎯Latency target table
| Workload | Good RTT | Watch above | Why it matters |
|---|---|---|---|
| Home lab storage | Under 2 ms | 5 ms | Small random IO feels slower |
| Remote desktop | Under 40 ms | 80 ms | Pointer and typing delay becomes obvious |
| Gaming | Under 60 ms | 100 ms | Input timing and interpolation suffer |
| VoIP / video calls | Under 150 ms | 200 ms | Conversation turn-taking degrades |
| Backup replication | Under 100 ms | 250 ms | TCP window sizing becomes important |
| Satellite failover | Under 700 ms | 900 ms | Interactive apps need tuning |
Latency exist between the user and every action performed within a lab. The longer it take for the data to travel from one point to another in the networks, the more higher the latency. This is due to the time that pass between the user initiating the command and the device receive the response.
Many people tend to notice the effect of high latency but not the reasons for it’s existence. However, latency comprise a few different component that sum to a total number for that specific task and network. The calculator will mathematically calculate the latency once you enter the details for you’re network path.
What Is Network Latency and How the Calculator Helps
It will save you from the work of having to calculate the latency yourself. Understanding why each parameter for the calculator are required will help you to better use the tool. The distance that the data must cover will establish the minimum latency for that path.
However, the medium in which the data travels will alter the speed at which it can travel. Fiber optic data travels faster than copper data by a few microsecond per kilometer. The number of bits in your data packet will impact the amount of time it take for the entire packet to be pushed across the network at the bottleneck speed.
A packet of data for gaming will travel differently than a data packet for backing up your computers. Jitter budget will impact your data traveling between your network device. It is important to note that while the latency may be high between two point with a stable latency, it can be even higher for real-time application if the latency fluctuate between two different value.
The calculator do not take into account human error or environmental change. An assessment of network latency made during quiet hour may show different result to when many people in the house are actively uploading data to the internet. While the calculator assume that the network latency is steady, actual networks are rarely without change.
Furthermore, the difference between the latency modeled in the tool and the actual latency that is measured on the network likely indicate a bufferbloat issue on the part of the ISP or WiFi network. Many people do not understand the balance between serialization and propagation latency. On short distance with high bandwidth, most of the data will be spent serializing the data into packets.
For data that must traverse long distances at slow speed, the data distance will have the highest latency, and data size will have little impact on the travel time. The tool make this tradeoff visible to the user. However, no memorization of latency formula is required.
By understanding the component that compose the latency of the network path, you can determine if upgrading the bandwidth will help with the latency issues. Another important latency component to discuss is jitter. For task like storage replication, jitter will have little impact on your network.
However, for task like video conferencing and remote desktop access, jitter will have a direct impact on the experience. The jitter margin will tell you how much latency jitter there is before the real-time application start to suffer from the latency in the network. The reference table provide typical latency for each component of the network.
For example, metro fiber distance will typically have latencies of single digit in milliseconds, whereas GEO satellites will have latencies of approximately 0.5 second. These typical latencies will help you understand your own latency measurement. Some of the common mistake people make with latency include treating it as a single value.
For instance, a high latency measurement may be blamed on distance, but it might actualy be the result of an overloaded router or ISP network. Furthermore, many people will attempt to reduce the latency of their network by shortening the length of the data cable between devices. However, the wireless latency between the computer and the first switch could contribute more to the latency than all of the data cables in the home network.
The calculator will help by enabling people to modify only one component at a time to see its impact on the network. The bandwidth-delay product help to show the relationship between latency and data throughput. For high-throughput networks with high latency, the bandwidth-delay product will be high.
This can create issue for TCP window sizing for data transmission. Modifying this will likely produce more significant results than attempting to reduce the latency of the network by a few millisecond. The goal of the tool is to provide a network with an understanding of whether or not the current network path can support the specific work that is to be perform on the network.
Tasks like storage cluster require the round trip to be below 2 millisecond. Gaming will have higher requirement for the average latency. Backups, however, are more lenient and will have higher allowed latency.
While the calculator will provide the breakdown of the network latency, it is up to the user to make a decision based on the number provided.
