Your ISP advertises a 100 Mbps plan. You download a file and your browser says 12 MB/s. Are you being throttled? No — both numbers are correct. They're just measuring different things, and the difference is exactly 8×.

Bits vs bytes

Bit (lowercase b): the smallest unit of data, a 0 or 1.

Byte (uppercase B): 8 bits. The standard unit for measuring file sizes.

Network speeds are advertised in bits per second (Mbps, Gbps). File sizes and download speeds use bytes per second (MB/s, GB/s). Convert by dividing by 8.

Why the dual-unit confusion exists

The convention dates back decades. Networking equipment historically measured throughput in bits per second because the equipment counted individual bit transitions on the wire. File systems measured in bytes because that's the natural unit of stored data.

The two industries kept their conventions. Now consumers see both — bits/s for advertised speed, bytes/s for actual throughput.

It also helps marketing: "1000 Mbps" sounds bigger than "125 MB/s." Same speed, different presentation.

The conversion formula

Divide Mbps by 8 to get MB/s.

  • 50 Mbps ÷ 8 = 6.25 MB/s
  • 100 Mbps ÷ 8 = 12.5 MB/s
  • 300 Mbps ÷ 8 = 37.5 MB/s
  • 1000 Mbps (1 Gbps) ÷ 8 = 125 MB/s

Multiply MB/s by 8 to get Mbps. This works for KB/KBps and GB/GBps too.

Worked example: download time

You're downloading a 10 GB game on a 100 Mbps connection.

  • Convert speed: 100 Mbps / 8 = 12.5 MB/s.
  • Download time: 10 GB / 12.5 MB/s = 10,240 MB / 12.5 MB/s = 819 seconds ≈ 14 minutes.

If your download manager shows 8 MB/s, you're at ~64 Mbps real throughput — about 64% of advertised speed. Real-world overhead (TCP, encryption, ISP throttling) typically eats 10–30%.

Common conversion errors

Confusing Mb and MB. 100 Mb (100 megabits) = 12.5 MB (12.5 megabytes). The lowercase/uppercase b/B is the only difference.

Treating 1 MB as 1 Mb in calculations. Off by 8×.

Comparing wired to wireless. A 1 Gbps wired connection vs 600 Mbps Wi-Fi (real throughput) — both are fast, but the underlying tech differs.

What you'd actually measure

Online speed tests (Fast.com, Speedtest.net) measure throughput in bits per second. Most show Mbps results.

Browser download progress meters show MB/s for files. Same data, different unit.

If you're downloading a file and see 12 MB/s, your connection is running at 96 Mbps (close to a 100 Mbps plan, accounting for ~4% overhead).

Network bottlenecks

Your end-to-end speed is limited by the slowest link. A 1 Gbps internet plan with an 802.11ac (Wi-Fi 5) router gets ~400–600 Mbps real Wi-Fi throughput. The router is the bottleneck.

Common bottlenecks:

  • Router Wi-Fi standard (older = slower)
  • Distance from router
  • Number of connected devices
  • ISP backbone congestion at peak times
  • Source server (downloading from a slow CDN)

Run speed tests at different times to identify which.

Is "Gigabit Internet" worth it?

1 Gbps = 125 MB/s real throughput. For typical household use:

  • Stream 4K to 5 simultaneous TVs (5 × 25 Mbps = 125 Mbps): 12.5% of capacity.
  • Multiple cloud backups + downloads + Zoom: maybe 100 Mbps total.
  • Most homes use under 200 Mbps even at peak.

Gigabit makes sense for: heavy file uploaders (content creators), households with many simultaneous heavy users, or future-proofing for unknown future needs.

For typical streaming households, 200–300 Mbps is plenty.

The "shared bandwidth" reality

Cable internet shares bandwidth with neighbors. A "300 Mbps plan" might deliver 250 Mbps at noon, 100 Mbps at 8 PM when everyone's streaming Netflix. Fiber doesn't share — you get the advertised rate any time of day.

This is why people pay more for fiber even when speeds look similar — guaranteed throughput vs best-case throughput.

Use the calculator

Our bandwidth calculator handles the conversion between Mbps and MB/s, plus download time estimates for any file size. Useful for pre-flight checks ("will this 50 GB game finish before bed?") or comparing internet plans.