Why a 1 TB drive shows up as 931 GB

Because the drive and the operating system are using two different definitions and only one of them says so. The manufacturer sells a terabyte as 1,000,000,000,000 bytes — the decimal meaning, which is what the prefix “tera” has meant since long before computers. Windows then divides that by 1,024 three times, gets 931.32, and labels the result “GB”. It is not gigabytes; it is gibibytes, and the wrong name has been printed. Nothing is missing from the drive, nothing is reserved and nothing is faulty: you have every one of the 1,000,000,000,000 bytes you paid for, measured with a ruler whose markings are 7.4% further apart. The same drive reports as 1 TB on macOS, which counts in decimal.

A kilobyte is 1,000 bytes. A kibibyte is 1,024. The binary prefixes — kibi, mebi, gibi and tebi, with the KiB, MiB, GiB and TiB symbols that go with them — were standardised by the IEC in 1998 to end exactly this ambiguity, because computers address memory in powers of two while “kilo” has meant a thousand for two centuries. The gap compounds at every step: a kibibyte is 2.4% more than a kilobyte, a mebibyte 4.9% more than a megabyte, a gibibyte 7.4% more, a tebibyte 10.0% and a pebibyte 12.6%. That is why the confusion is invisible on a text file and glaring on a hard drive.

Because a megabit is not a megabyte — it is one eighth of one. A 100 Mbps connection carries 100,000,000 bits every second, and eight bits make a byte, so that is 12,500,000 bytes, or 12.5 MB per second. A 100 MB file therefore takes eight seconds at the full line rate, before any protocol overhead or sharing. You can check that on the calculator above: enter 100,000,000 and convert bit to MB. Connections are advertised in bits because the numbers come out eight times larger; browsers report downloads in bytes because that is how files are measured.

It depends which computer. Windows File Explorer divides by 1,024 and writes KB, MB and GB, so its figures are really kibibytes, mebibytes and gibibytes wearing decimal names. macOS has counted in decimal since Snow Leopard, so a file it calls 1 GB really does hold 1,000,000,000 bytes. On Linux, df and ls default to 1,024 with single-letter suffixes and take an --si flag to switch to 1,000. This is why the same file can be reported at two different sizes on two machines without either being broken, and why a size quoted with no context is worth converting rather than trusting.

Because memory has to be addressed in binary and storage does not. A memory chip is wired so that every address is a fixed number of bits, which forces the total capacity to a power of two — so RAM arrives as 8, 16, 32 and 64, and “8 GB of RAM” genuinely means 8 GiB, which is 8,589,934,592 bytes. A drive has no such constraint; its capacity is whatever the controller reports, so manufacturers choose round decimal numbers that read well on a box. Both conventions live inside the same machine, which is a large part of why this confusion has never gone away.

Because they are nearly always used as rates rather than amounts — Mbps is megabits per second, and a rate needs a time unit that an amount converter has no business asking for. The bit is in the list for the cases where you genuinely hold a count of bits, and a megabit is simply a million of them, so entering 100,000,000 covers a 100 Mbps figure directly. Adding six more entries that are only ever half of a measurement would make the list harder to scan in exchange for a multiplication anyone can do.

No. Everything is worked out in your own browser, there is no server behind this site, and this page writes nothing to your device. Close the tab and every figure is gone.