Guide
100 Mbps does not mean 100 MB/s. That tiny “b” is an 8× difference.
If a 1 GB transfer estimate looks suspiciously fast, check the units first. Network speeds are usually bits; file sizes are usually bytes.
100 Mbps is 12.5 MB/s, not 100 MB/s
Network speed is commonly written in Mbps — megabits per second. File size is commonly written in MB or GB — megabytes. One byte is eight bits.
So 100 Mbps is 12.5 MB/s in theory. A 1 GB file takes roughly 80 seconds at that rate before any overhead. Treating 100 Mbps as 100 MB/s produces a neat-looking answer that is eight times too optimistic.
The math can be right and the transfer can still be slower
Your advertised link rate is a ceiling, not a guaranteed file-copy rate. Protocol overhead, encryption, Wi‑Fi, VPNs, storage speed, and the remote server all take a bite out of throughput.
If a real test averages 8 MB/s, use 8 MB/s for the practical estimate. Keeping both the theoretical minimum and the measured estimate makes the gap much easier to explain.
One big file and a million tiny files are different jobs
Two transfers can have the same total size and finish at very different times. Tiny files add repeated work: metadata, file creation, request/response cycles, and other per-file overhead.
When total size looks harmless but the job drags on, check the file count too. Bundling many tiny files into an archive can sometimes change the result dramatically.
A two-pass estimate works well: theory, then measurement
First normalize the units and calculate the best-case time. Then run a representative transfer and replace the theoretical speed with measured MB/s.
For a migration or deadline-sensitive upload, add headroom for busy periods and slowdowns. That is usually more useful than pretending the advertised Mbps number is the whole story.