Bandwidth Converter

What this tool is used for

• Working out how long a transfer of a given size will take at a quoted line speed.
• Converting a link speed in megabits into the megabytes per second a download shows.
• Comparing two connection specifications quoted in different units.
• Sizing a link against an expected throughput requirement.
• Checking whether a measured speed matches what was sold.

Frequently Asked Questions

Why is my 100 Mbps connection only 12 MB/s?

Because a byte is eight bits, and network speeds are quoted in bits while file sizes are in bytes. 100 Mbps ÷ 8 = 12.5 MB/s before any overhead. A lower-case b and an upper-case B are the only thing distinguishing them, which is why ISPs are comfortable quoting the bigger number.

Does this use 1000 or 1024 as the multiplier?

1000. Network units are decimal SI throughout — a kilobit per second is exactly 1000 bits per second, not 1024 — unlike RAM, and unlike the way file managers historically reported disk sizes. Mixing the conventions introduces about 2.4% of error per prefix.

How do I work out a download time?

Convert the file size to bits (bytes × 8) and divide by the connection's bits per second. A 5 GB file on 100 Mbps is 40,000 megabits ÷ 100 = 400 seconds, about seven minutes at theoretical maximum — real transfers land 10–20% below once TCP and protocol headers are counted.

Why is real throughput always below the rated speed?

Because the rating is the physical line rate and your payload shares it with headers. Ethernet, IP and TCP each add framing, TCP acknowledges and retransmits, and Wi-Fi adds contention. 90% of line rate on a wired link is good; wireless routinely delivers half its advertised figure.

Is 5G actually faster than a 1 Gbps fibre line?

In peak lab conditions, sometimes. The benchmarks here — 300 Mbps typical for 5G against 1000 for fibre — reflect that radio is shared and distance-sensitive while a fibre line is dedicated. Latency differs by more than throughput does, and latency is what interactive work notices.

Common errors and gotchas

• Confusing bits with bytes, a factor of eight and by far the most common error here.
• Assuming a line speed equals achievable throughput, which protocol overhead reduces.
• Mixing decimal and binary prefixes, where a megabyte may mean 10^6 or 2^20 depending on context.
• Reading an upload figure as symmetrical, when most consumer links are not.
• Comparing a peak rate against a sustained one as if they were the same measurement.