Skip to content

MB vs MiB (Decimal vs Binary Sizes)

Megabytes versus mebibytes—compare decimal and binary size units and why disk sizes never match the label.

Overview

The gap between decimal and binary size units is why a “1 TB” drive shows up as roughly 931 GB. Vendors count in powers of ten; many operating systems count in powers of two and still print the decimal abbreviation. The units are both correct, but mixing them in one calculation produces errors that grow with scale—about 7% per prefix step.

Pros

  • Matches SI prefixes users expect
  • Used by drive makers and bandwidth specs
  • Simple powers of ten arithmetic

Cons

  • Does not align with memory addressing
  • Looks larger than the operating system reports

Pros

  • Aligns with powers of two in memory
  • Unambiguous in technical documentation
  • Matches many operating system readouts

Cons

  • Unfamiliar to non-technical readers
  • Rarely used in consumer marketing

Comparison table

AspectMB (decimal)MiB (binary)
Definition1 MB = 1,000,000 bytes1 MiB = 1,048,576 bytes
Who uses itDrive and network vendorsOperating systems and specs
Drift at gigabyte scaleBaselineAbout 7% larger per unit
Best fitYou quote storage, bandwidth, or marketing figuresYou document memory, buffers, or file system limits

Recommendation

Pick one convention per document and use the explicit prefix: MB for decimal, MiB for binary. When a limit really is a power of two, say MiB so nobody re-derives it incorrectly.

Related tools

Related articles

Frequently asked questions

Which unit do upload limits use?
It depends on the implementation. If a limit is enforced as 10,485,760 bytes it is 10 MiB even when the message says 10 MB—check the byte count rather than the label.
Should user-facing text say MiB?
Usually no. Show MB to end users and reserve MiB for technical documentation where the exact power-of-two boundary matters.
What pushes someone toward MB (decimal) instead of MiB (binary)?
MB (decimal) wins when you quote storage, bandwidth, or marketing figures. The practical upside is that matches SI prefixes users expect, and used by drive makers and bandwidth specs. The trade-off to watch is that does not align with memory addressing.
When does MiB (binary) beat MB (decimal) for the same job?
Reach for MiB (binary) when you document memory, buffers, or file system limits. It gives you aligns with powers of two in memory plus unambiguous in technical documentation, at the cost that unfamiliar to non-technical readers.
Can ToolHub help me try MB (decimal) and MiB (binary) before I commit?
Yes. The tools linked from each side of MB vs MiB (Decimal vs Binary Sizes) run in your browser, so you can exercise MB (decimal) and MiB (binary) with sample data without uploading anything.