Why your 1 TB drive shows 931 GB: decimal vs. binary storage units

Two meanings of “giga”

In the SI system, the prefix giga means 10⁹ — exactly one billion. Storage manufacturers use this definition, so 1 TB = 1,000,000,000,000 bytes. Computers, however, address memory in powers of two, and software has historically used “kilobyte” to mean 1,024 bytes, “megabyte” 1,048,576 bytes, and “gigabyte” 1,073,741,824 bytes.

To end the ambiguity, the IEC introduced binary prefixes in 1998: kibibyte (KiB), mebibyte (MiB), gibibyte (GiB), tebibyte (TiB). NIST recommends them. Windows still labels binary values with the decimal names — it says “GB” but computes GiB. macOS switched to decimal units in 2009, which is why the same drive shows 1 TB on a Mac and 931 GB on a PC.

The maths

1 TB = 10¹² bytes. One GiB = 2³⁰ = 1,073,741,824 bytes. 10¹² ÷ 1,073,741,824 = 931.32 GiB. The ratio is the same for every size: divide decimal gigabytes by 1.0737 to get binary gibibytes, or by 1.0995 to convert decimal terabytes to tebibytes.

Where the rest goes

After the unit difference, a small amount of real capacity is used by the file system’s own structures (typically under 1%) and, on system drives, by recovery partitions and hibernation files. SSDs also reserve a few percent internally for wear-levelling, but that is already excluded from the advertised figure.

RAM is the exception

Memory chips are built in powers of two, so RAM is always quoted in binary units even when the label says GB: a “16 GB” memory kit is 16 GiB = 17,179,869,184 bytes. This is why your operating system reports RAM at its full advertised figure but storage at roughly 93%. Flash storage in phones and SSDs follows the storage convention (decimal), which is why a 128 GB phone shows about 119 GB before the operating system takes its share.

How to check the real capacity yourself

• Windows: right-click the drive → Properties. The “Capacity” line shows the exact byte count next to the GB figure. Divide bytes by 1,000,000,000 to confirm the advertised decimal size.
• macOS: Finder → Get Info shows decimal GB with the byte count in brackets; Disk Utility does the same.
• Linux: `df -H` prints decimal units, `df -h` prints binary units — same drive, two answers.
• Phones: Settings → Storage rounds to the nearest whole unit and includes the system partition, so expect the largest apparent “loss”.

Download speeds: bits, not bytes

Internet plans are sold in megabits per second (Mb/s); files are measured in megabytes (MB). Eight bits make a byte, so a 100 Mb/s connection moves at most 12.5 MB/s. A 4.7 GB film therefore takes at least 4,700 ÷ 12.5 ≈ 376 seconds, about six and a half minutes, in ideal conditions. Real transfers run 10–20% slower because of protocol overhead, so treat the figure as a floor.

Does the difference matter in practice?

For a single drive, rarely — 7% is a rounding error against how quickly photos and video fill space. It matters when you plan backups or RAID arrays, where mixing a decimal spreadsheet with binary system reports can leave you short by a whole drive’s worth over a large pool; when comparing cloud storage plans (almost all decimal) with local usage (often binary); and when a drive appears “faulty” to a customer who expected 1,000 GB. In every case the fix is the same: pick one unit system, write it down, and convert with the 1.0737 factor.