What is network throughput?
Network throughput is the amount of useful data successfully delivered over a network connection in a given amount of time. It is typically measured in bits per second (bps), megabits per second (Mbps), or gigabits per second (Gbps). In practice, throughput is almost always lower than the raw link rate because real traffic includes packet headers, acknowledgments, retransmissions, contention, and other overhead.
Throughput vs. bandwidth vs. speed test results
Bandwidth (capacity)
Bandwidth is the theoretical maximum rate your connection can carry, such as a 1 Gbps Ethernet link. Think of it as the width of a pipe.
Throughput (delivered data)
Throughput is what actually gets through the pipe in normal operation. This is what users experience when transferring files, streaming video, backing up data, or syncing cloud storage.
Goodput (application payload only)
Goodput is payload data only—excluding protocol headers and control traffic. It is often the most realistic measure for user-facing performance.
Formula used by this calculator
The calculator uses a practical approximation:
- Effective Throughput = Link Bandwidth × (1 − Overhead%) × Utilization% × (1 − Loss%)
- Transfer Time = Data Size (in bits) ÷ Effective Throughput (in bits/sec)
- BDP (optional) = Effective Throughput × RTT
This approach helps with planning and estimation, especially when comparing links, sizing transfer windows, or forecasting backup duration.
How to interpret your results
Raw throughput
The raw value reflects the nominal line rate before losses and inefficiencies.
Effective throughput
Effective throughput is a more realistic number for real-world operations. If it is much lower than expected, check congestion, packet loss, duplex mismatches, Wi-Fi interference, and TCP tuning.
Estimated transfer time
This estimate is useful for maintenance windows, migration tasks, software distribution, and large data replication jobs.
Typical factors that reduce throughput
- Protocol overhead (Ethernet/IP/TCP/UDP/TLS/VPN encapsulation)
- Latency and window size limitations
- Packet loss and retransmissions
- Jitter and queueing under congestion
- Shared medium contention (especially Wi-Fi)
- Endpoint bottlenecks (CPU, disk I/O, NIC offload settings)
Ways to improve throughput
- Use wired links where possible for predictable performance
- Reduce packet loss by fixing physical layer issues and queue drops
- Enable modern congestion control and proper TCP window scaling
- Minimize unnecessary encapsulation layers
- Apply QoS policies for critical traffic
- Upgrade slow storage paths that cap transfer rates
Example scenario
Suppose you have a 500 Mbps WAN link, 10% overhead, 80% utilization, and need to move 250 GB of data. Your effective throughput is approximately 360 Mbps. That is about 45 MB/s, so transfer time is roughly 1 hour and 35 minutes. This quick estimate can prevent under-sized migration windows and failed overnight jobs.
Final note
Use this calculator for planning and first-pass sizing. For production validation, combine these estimates with real measurements from tools such as iPerf, interface counters, flow telemetry, and application-level transfer logs.