Data Storage Converter

Type a size, choose the unit it is in, and pick the unit you want. The converter handles both families of storage units: the decimal ones printed on drive boxes and spec sheets (KB, MB, GB, TB) and the binary ones your operating system may actually count in (KiB, MiB, GiB, TiB). Plain bytes sit in the middle as the common anchor, so the answer comes out exact whenever bytes are one end of the conversion and rounded to four decimals otherwise.

How the conversion works

Every unit is a fixed number of bytes, so any conversion is a multiply followed by a divide:

result = value × bytes_from ÷ bytes_to

The defaults show the single most-searched case. 512 GB is 512 × 10⁹ = 512,000,000,000 bytes. One GiB is 2³⁰ = 1,073,741,824 bytes, and 512,000,000,000 ÷ 1,073,741,824 = 476.8372 GiB. That one division is the entire mystery of the shrinking SSD. The decimal units are exact powers of 1,000 and the binary units exact powers of 1,024, so the conversion factor itself is never approximate; only the four-decimal display rounds.

The byte and the bit

The smaller unit came first. John Tukey, a mathematician at Bell Labs, contracted "binary digit" to "bit" in a memo written in 1947, and Claude Shannon put the word into print in his 1948 paper "A Mathematical Theory of Communication", the founding document of information theory. A bit is a single yes-or-no, a one or a zero, and everything else is a count of them.

The byte was named nine years later. Werner Buchholz coined it at IBM in June 1956 while the company was designing the 7030 Stretch, an early transistorized supercomputer. He respelled "bite" as "byte" so it could not be misread as "bit". In that first design a byte was not fixed at eight bits; it could be any width from one to six, sized to whatever the input-output hardware happened to need. The eight-bit byte became standard when IBM announced the System/360 in April 1964, a family of machines built to run the same software across every model. Fred Brooks, who managed the System/360 project under chief architect Gene Amdahl, later called changing the series from a six-bit byte to an eight-bit byte the most important single decision he ever made, because eight bits left room for lowercase letters, and the choice then propagated to nearly every computer built afterward.

Storage in that era was counted in a handful of characters. IBM's Model 350, the disk unit inside the RAMAC announced in September 1956 and the first commercial hard drive, held about five million six-bit characters, or 3.75 MB, across fifty aluminium platters two feet across turning at 1,200 rpm in a cabinet the size of two refrigerators. The complete 305 RAMAC system it belonged to leased for roughly 3,200 dollars a month. A microSD card the size of a fingernail now holds hundreds of thousands of times as much, which is why the units in this converter reach up to the terabyte.

Where the 1,024 habit came from

Memory chips are addressed in powers of two, so a given number of address lines maps naturally onto capacities of 1,024, 65,536, 1,048,576 and upward. Early programmers borrowed the metric prefix kilo for 1,024 because it was close enough — only 2.4 percent off — and the shortcut stuck. The trouble is that the error compounds with every step up the scale:

The names began with a 1995 proposal from an IUPAC nomenclature committee, which suggested kilobinary, megabinary and gigabinary clipped to kibi, mebi and gibi. The International Electrotechnical Commission took the idea up to end the ambiguity, circulated a draft in 1996, approved the result in December 1998, and published it in January 1999 as Amendment 2 to the IEC 60027-2 standard, covering kibi, mebi, gibi, tebi, pebi and exbi with the two-letter symbols Ki, Mi, Gi, Ti, Pi and Ei. The IEEE raised the same prefixes to a full-use standard, IEEE 1541, in March 2005. The prefixes are exact and unambiguous, but adoption stayed patchy. RAM is still sold in true binary sizes under plain decimal labels — a 16 GB memory module actually holds 17,179,869,184 bytes, which is 16 GiB — because memory is manufactured in powers of two and the JEDEC standards that govern it kept the older usage. Hard drives, SSDs and SD cards, by contrast, are genuinely decimal.

Decimal marketing, binary reporting

A drive sold as 512 GB really does hold 512 billion bytes. The number looks smaller in Windows because File Explorer counts in binary units while printing decimal labels, so the same drive appears as 476.84 "GB". Neither figure is wrong; they are the same quantity of bytes measured with two different rulers. Apple removed the discrepancy on the Mac in 2009, when Mac OS X 10.6 Snow Leopard switched to decimal reporting so a drive shows the same number the manufacturer prints on the box. Linux desktops split: GNOME's Files has shown decimal sizes since 2011, while KDE's Dolphin counts in binary and at least labels it honestly as KiB, MiB and GiB.

The gap was contentious enough to reach the courts. Western Digital settled a United States class action in 2006, filed the previous year over an 80 GB drive that reported about 74.4 gigabytes to the operating system, roughly seven percent short of the number on the box. The company gave about a million buyers a free copy of backup software and agreed to carry a capacity disclaimer. Seagate settled a parallel case in 2007 with a choice of free software or a cash refund of five percent of the purchase price, and agreed to spell the point out in its packaging: one gigabyte means one billion bytes when referring to hard drive capacity, an operating system may use a different standard and report a lower capacity, and formatting takes some of what is left. A terser footnote along the lines of 1 GB equals 1,000,000,000 bytes was already on drive boxes by the early 2000s, before either case was filed. What the settlements bought was the fuller explanation rather than the disclosure itself.

Which convention applies where

This is one unit dispute that does not split along national lines. A drive box in the US, UK, EU, Canada or Australia uses the same decimal gigabyte, and Windows shows the same binary count in every region. The split runs by industry instead: storage marketing and network speeds are decimal, RAM and many programming contexts are binary, and operating systems disagree with one another. When two tools report different sizes for the same file, the first thing to check is which family each one is counting in.

What the number does and does not include

The converter changes units only. It treats the size you enter as raw capacity and subtracts nothing for the filesystem, so a freshly formatted drive shows slightly less usable space than the pure conversion suggests. Formatting reserves room for the file table and journal, and a factory drive often carries a hidden recovery partition as well, which together can remove a further one to two percent on top of the decimal-to-binary label gap. When you need the space actually available to copy files into, read it from the operating system after formatting rather than from the box.

Frequently asked questions

Why does my 512 GB SSD show only 476 GB in Windows?

The drive contains the full 512,000,000,000 bytes printed on the box. Windows divides by 1,073,741,824 — one gibibyte — but labels the result GB, which gives 476.84. The label mismatch removes nothing, though a recovery partition and filesystem overhead can shave off a little more on top of it.

What is the difference between a kilobyte and a kibibyte?

A kilobyte (KB) is 1,000 bytes under the SI definition, while a kibibyte (KiB) is 1,024 bytes. The IEC introduced the kibi prefix in 1998 to end decades of ambiguity in which KB could mean either number depending on context. The gap is only 2.4 percent at this scale but compounds to almost 10 percent by the terabyte level.

Is a terabyte 1,000 GB or 1,024 GB?

A terabyte is 1,000 GB by definition, and that is what drive makers ship. The 1,024 relationship belongs to the binary family: one tebibyte (TiB) is 1,024 GiB, or about 1.0995 TB. A drive sold as 1 TB shows roughly 931 GiB in Windows, which reports binary counts under decimal names.

How many megabytes are in a gigabyte?

Exactly 1,000 MB in a GB within the decimal system used on storage packaging. In the binary family, one GiB is 1,024 MiB, and converting across families gives one GiB equal to about 1,073.74 MB. Mixing the two families is where most of the confusion starts, so check which one your software reports.