Data Storage Converter

Why a 1 TB drive shows as 931 GB

This is the most common question about data units, and the answer is not that anyone is cheating. A drive sold as 1 TB holds exactly one trillion bytes — 10¹² — which is what the manufacturer promised, and the SI has meant 10¹²by “tera” since long before hard drives existed.

Your operating system then divides that by 2³⁰ — 1,073,741,824 — because it counts in powers of two, and writes the result on screen with a decimal label. One trillion divided by 1,073,741,824 is 931.32, and the screen says “931 GB”.

So the drive contains exactly what it says. The 931 is the same quantity measured with a bigger unit that has been given the smaller unit’s name. Nothing is lost between the two figures — they are the identical number of bytes.

The gap widens at every step, because it compounds: a kilobyte is 2.4% off, a megabyte 4.9%, a gigabyte 7.4%, and a terabyte 9.9%. That is why the effect is barely noticeable on a memory card and obvious on a multi-terabyte drive.

Two prefix systems, one vocabulary

Decimal (SI) prefixes are powers of a thousand, exactly as they are everywhere else in measurement: a kilobyte is 1,000 bytes, a megabyte 10⁶, a gigabyte 10⁹. Nothing about data makes kilo mean anything else.

Binary (IEC) prefixes are powers of 1,024, and have their own names precisely so the decimal ones can keep theirs: a kibibyte (KiB) is 2¹⁰bytes, a mebibyte (MiB) 2²⁰, a gibibyte (GiB) 2³⁰. They were standardised in 1998 for exactly this problem.

Both are exact and both are correct. The confusion is entirely a labelling habit: computing adopted the SI words for binary quantities decades before the IEC names existed, because 1,024 is close enough to 1,000 that borrowing “kilo” seemed harmless — and at that scale it nearly was. It stopped being harmless as the numbers grew, since the error compounds with every prefix.

This converter lists both families with their real names and never presents one as the true meaning of the other. If your source says “GB” and you need to know which it meant, the next section is the shortcut.

Bits and bytes are not the same

A second eight-fold trap sits underneath the first. A bit is a single binary digit; a byte is eight of them. Storage is measured in bytes and written with a capital B; network and connection speeds are measured in bits and written with a lower-case b.

So a 100 Mbps connection is 100 megabits per second, which is 12.5 megabytes per second — divide by eight. A 1 GB file over that link takes about 80 seconds at best, not 10. Internet providers quote in bits because the number is eight times larger, and both conventions are legitimate; the capital letter is the only thing telling them apart, and it is easy to miss.

Combined with the prefix problem, the two effects compound in the same direction: a download shown at “12.5 MB/s” by a browser using binary megabytes is moving about 13.1 decimal megabytes per second. That is why a transfer estimate and the arithmetic you did in your head rarely agree.

What your operating system actually reports

The systems genuinely disagree, so the same drive shows different numbers depending on what you plug it into — and neither machine is wrong.

Windows computes in binary and labels decimally: it divides by 2³⁰and writes “GB”. This is the combination that produces the 931 figure. macOS has reported decimal units since OS X 10.6, so it shows a 1 TB drive as about 1,000 GB and agrees with the box. Linux varies by tool: df -h uses binary by default while df -H uses decimal, and most desktop file managers follow the IEC names properly.

RAM is the consistent exception. Memory really is manufactured in powers of two, because addressing is binary, so a 16 GB module genuinely holds 2³⁴ bytes — 16 GiB. There is no discrepancy to explain there, which is why memory sizes always look like round numbers and drive sizes do not.

One further reduction is separate from all of this: formatted capacity is lower than raw capacity, because the filesystem needs space for its own structures. That is a genuine loss of usable space, unlike the prefix difference, and it depends on the filesystem and the block size.

Doing it in a spreadsheet

The factors are worth writing out, because there is no unit code that distinguishes the two systems: =A1/1024 for KiB to MiB, =A1/1000 for kB to MB, and =A1*1000000000000/1073741824 for TB to GiB — the formula behind the 931.

For the bits-and-bytes step, keep the division by eight as its own visible cell rather than folding it into a rate: =A1/8 turns megabits per second into megabytes per second, and a sheet that hides it is a sheet where a transfer estimate is eight times wrong with nothing on screen to show why.

If you are building a capacity plan, pick one system for the whole sheet and put its name in the column header. Mixing them silently is exactly how the 931 problem arises in the first place, and a header costs nothing.

Related calculators

Where a data figure usually goes next:

More in Conversion, or browse all calculators.

Sources and methodology

Nothing here is fetched and there is no data feed: every figure is computed from what you type, on factors that are exact in both systems. The sources settle the two definitions and, importantly, the fact that operating systems genuinely disagree with each other — which is why the same drive shows one number on a Mac and a different one on Windows.

• IEC 80000-13 — binary prefixes kibi, mebi, gibi and the units KiB, MiB, GiBInternational Electrotechnical Commission / ISO · verified 2026-09-06 · The binary prefixes as exact powers of 1 024, standardised precisely so that the decimal SI prefixes could keep their own meaning
• The International System of Units (SI), 9th edition — the SI prefixes as powers of tenBureau International des Poids et Mesures · verified 2026-09-06 · That kilo, mega, giga and tera are defined as 10³, 10⁶, 10⁹ and 10¹² in every context, including data
• NIST reference on prefixes for binary multiplesNational Institute of Standards and Technology · verified 2026-09-06 · The recommendation that binary multiples use the IEC prefixes and that SI prefixes retain their decimal meaning — the position this page follows
• macOS reports storage in decimal units since OS X 10.6Apple Support · verified 2026-09-06 · That different operating systems genuinely disagree — macOS reports decimal GB while Windows reports binary quantities labelled GB, so the same drive shows two different numbers