Professional scenario conversions
Mb vs MB in data transfer specs
Megabits per second describe link capacity; megabytes per second describe file throughput. A factor-of-eight gap plus protocol overhead confuses SLAs and user expectations.
Goodput—what file copy shows—is always lower than PHY headline Mbps because of headers, retransmits, and congestion control.
Key takeaways
- Divide Mbps by 8 for theoretical MB/s; expect lower real throughput.
- Quote both peak Mbps and p95 RTT when user experience matters.
- Wi-Fi headline ≠ TCP goodput—test with your workload.
- MiB vs MB: storage often binary; networks decimal—label both in docs.
How to convert
1 B = 8 b
Bits for networks, bytes for files
ISPs advertise Mbps because serial links count bits; OS file copy dialogs often show MB/s.
The factor of eight
Divide Mbps by 8 to get theoretical megabytes per second before overhead; real throughput is lower.
MiB vs MB ambiguity
Storage vendors sometimes use binary megabytes; network rates are decimal megabits—do not mix silently.
Latency is not bandwidth
A high Mbps plan does not fix high round-trip time; user experience needs both metrics.
Writing requirements
Specify “100 Mbps symmetric” vs “12.5 MB/s observed file upload” explicitly when testing WAN links.
PHY rate vs TCP goodput
Wi-Fi box Mbps is a layer-1 headline; file copies measure goodput after headers, retransmits, and congestion control.
Jitter and bufferbloat
High Mbps with unstable latency still hurts video calls—quote both peak Mbps and p95 RTT in SLAs when relevant.
Cellular NSA vs SA labeling
Carrier “5G” Mbps numbers may aggregate non-standalone anchors; compare field tests, not brochure peaks.