Free online calculator
Network Transfer Time Calculator
Free network transfer time calculator. Estimate backup, file copy, and media library transfer duration with realistic protocol overhead for Ethernet, Wi-Fi, and USB.
Why real transfers are slower than link speed
Ethernet speed is measured in bits per second, while file sizes are usually shown in bytes. Protocol overhead, storage speed, CPU load, file count, encryption, Wi-Fi conditions, and retransmissions reduce application throughput.
For a healthy wired link and large sequential files, 90-94% efficiency is a useful planning range. Small files, Wi-Fi, slow disks, SMB signing, and busy systems can be much lower.
Use the slower component
A 2.5GbE link cannot make a source drive read faster or a destination write faster. If storage throughput is known, convert it to network speed and use the lower limit.
Measured 1GbE example
Our EQi12 large-file SMB test sustained about 102.95–107.15MiB/s over Gigabit Ethernet with matching SHA-256 hashes. Entering a measured application rate is more accurate than assuming the nominal 1Gbps link. A continuous 1TiB transfer near 105MiB/s takes roughly 2.7 hours before additional filesystem or verification work.
The same machine moved 10,000 small files much more slowly because each file adds metadata, open/close, security, and protocol work. Use the network and SMB test to see why one large file and a directory tree need different planning margins.
Measured anchors from the EQi12 lab
Generic efficiency guesses are fine for a first pass, but a measured anchor beats any estimate. All values below are first-party measurements on a Beelink EQi12 (i3-1215U, 2×2.5GbE NICs negotiated at 1GbE for the wired tests), published in the open EQi12 Measurement Data repository (CC BY 4.0).
| Path | Measured result | Test condition |
|---|---|---|
| Wired 1GbE iperf (NIC1) | 890.75 / 902.57 Mbps | 4 GiB iperf, both directions, Jul 2026 |
| Wired 1GbE file copy | 108.76–112.2 MiB/s | 4 GiB single file, PowerShell copy, both directions |
| Wired 1GbE SMB large file | 102.95–107.15 MiB/s | SMB share, SHA-256 verified, Jul 2026 |
| Wi-Fi 6 iperf (near range) | 618.37 / 870.18 Mbps | 4 GiB iperf, negotiated 2402 Mbps, near range |
| USB fast port → SSD | 300.64–301.91 MiB/s write | 50 GiB sustained write, same drive and enclosure |
| USB slow port → SSD | 36.36 MiB/s write | 50 GiB sustained write, identical drive — 8x port lottery |
| Internal NVMe (WD SN540) | 623.29 MiB/s write / 468.2 MiB/s read | 50 GiB sustained, SHA-256 verified, Jul 2026 |
Planning rule that falls out of this table: convert your slowest storage number to a link equivalent and enter it as the speed with 100% efficiency, rather than pretending a gigabit link will run at line rate into a 36 MiB/s USB port.
Choose an efficiency value
| Workload | Starting estimate | What can lower it |
|---|---|---|
| Wired, large sequential file | 90–94% | Slow storage, CPU, encryption |
| Wi-Fi large file | 40–75% | Signal, shared airtime, interference |
| Many small files | Use a measured MB/s value | Metadata, antivirus, filesystem latency |
| Internet transfer | Use the slower WAN direction | ISP shaping, remote server, latency |
Decimal and binary units
This calculator uses decimal GB/TB and decimal Mbps/Gbps, matching network and drive marketing. If your file manager reports GiB or TiB, expect a unit difference. Always state the unit when publishing a result.