Digital Storage Conversion Guide

← Voltar ao conversor de Armazenamento Digital

Tradução em andamento.

Este guia ainda não foi traduzido para PT. A versão em inglês é mostrada abaixo.

Digital storage is the conversion category that ate its own foot. Hard-drive vendors sell capacity in decimal gigabytes — 1 GB = 1,000,000,000 bytes. Operating systems report capacity in binary gibibytes — 1 GiB = 1,073,741,824 bytes — but call them 'GB'. A nominal 1 TB drive arrives, you plug it in, and your computer says 931 GB free, which is exactly right and looks entirely wrong. This guide covers the decimal SI prefixes, the binary IEC prefixes, the units we support, and the specific places the gap shows up in real product specs.

What is digital storage?

Digital storage measures how much information a device holds, in bits or bytes. One byte is eight bits, by long-standing convention dating to the 1960s. Every other storage unit is a multiple of bytes. The complication is that there are two competing scaling conventions: decimal (powers of 1,000) and binary (powers of 1,024). Decimal is the SI standard and matches the metric prefixes used elsewhere. Binary is the historical convention from 1960s mainframes, where memory addressing made 1,024-byte blocks the natural unit.

Modern standards split the prefixes by family. SI prefixes (kilo, mega, giga, tera) stay decimal: 1 kB = 1,000 B. IEC prefixes (kibi, mebi, gibi, tebi) are explicitly binary: 1 KiB = 1,024 B. The IEC family was standardised in IEC 80000-13 (originally IEC 60027-2) and ratified by IEEE 1541-2021. Despite that, popular usage in operating systems and tooling still abuses 'KB' and 'MB' to mean the binary values.

History of digital storage units

Early computers stored everything in powers of two. IBM's System/360 in 1964 set a 1,024-byte page size, and the casual notation '1 kilobyte' meant 1,024 bytes. The decimal-prefix purists protested for decades — the SI definition of 'kilo' has always been 1,000 — but the binary usage stuck through the 1990s, especially in RAM specifications.

The hard-drive industry settled on decimal early. A '20 megabyte' drive in 1985 held 20,000,000 bytes, not 20,971,520. Network speeds followed: '100 Mbps' Ethernet is 100,000,000 bits per second, not 104,857,600. Memory (RAM) followed binary: a '512 MB' DIMM holds 512 × 1,048,576 = 536,870,912 bytes. The split has held for forty years, with the result that 'gigabyte' in a marketing context is decimal and 'gigabyte' in an OS context is usually binary.

IEC 80000-13 (originally published 1999, revised periodically) and IEEE 1541-2021 codified the kibi/mebi/gibi/tebi prefixes specifically to remove this ambiguity. Linux file utilities (`df`, `du`) added the new prefixes years ago. Apple's macOS and most modern Linux distributions switched their file-manager display to decimal GB in the 2010s. Microsoft Windows continues to label binary gibibytes as 'GB' as of 2026, which is the source of most 'why is my drive smaller than advertised' questions.

Units we cover

Our digital-storage category supports eleven units across two families. Decimal SI: bit, byte, kilobyte (1,000 B), megabyte (10⁶ B), gigabyte (10⁹ B), terabyte (10¹² B), petabyte (10¹⁵ B). Binary IEC: kibibyte (1,024 B), mebibyte (1,024² B), gibibyte (1,024³ B), tebibyte (1,024⁴ B). Each unit has its exact byte equivalent published, and Calcflux applies those factors directly.

We do not introduce '1024-byte kilobyte' as a separate unit — that is the historical conflation we are trying to disambiguate. When a spec or operating system displays 'GB' for what is technically GiB, the user has to decide which interpretation matches their context. Our converter accepts both families on equal footing so you can answer 'how many GiB is a 1 TB drive' without re-reading IEC 80000-13.

How to convert

Every digital-storage conversion routes through the byte. Each unit declares how many bytes it represents: 1 kB = 1,000, 1 MB = 1,000,000, 1 GB = 1,000,000,000, 1 KiB = 1,024, 1 MiB = 1,048,576, 1 GiB = 1,073,741,824, 1 TiB = 1,099,511,627,776. The methodology page lists every factor and cites IEC 80000-13.

Useful approximations: 1 GiB ≈ 1.0737 GB. 1 TiB ≈ 1.0995 TB. The gap grows with each prefix step: kilo is off by 2.4%, mega by 4.9%, giga by 7.4%, tera by 10.0%, peta by 12.6%. Hard-drive 'shrinkage' as displayed by Windows-style OSes is exactly this gap — a 1 TB drive ships with 1,000,000,000,000 bytes, which Windows divides by 1,073,741,824 to display 931.32 'GB' (really GiB).

Round-number capacity anchors are worth memorising. One byte holds a single ASCII character. One kilobyte holds about a page of plain text. One megabyte holds a minute of 128-kbps MP3 audio. One gigabyte holds about three hours of standard-definition video. One terabyte holds roughly 250 high-definition movies or 250,000 high-resolution photos. Petabyte-scale storage is the everyday unit for data-centre planning; a single cabinet today routinely holds 1 PB of usable capacity.

Common conversions you'll need

Storage planning and OS spec questions drive most digital-storage searches. Megabytes to gigabytes handles file-size estimates — a 4,800 MB game install is 4.8 GB on the box. Gigabytes to terabytes handles drive sizing and NAS planning — 4,000 GB is 4 TB of nominal capacity. Kilobytes to megabytes covers historical file sizes; bytes to megabytes covers programmatic outputs from `wc -c` and similar tools. For the decimal-vs-binary question specifically, mebibytes to megabytes is the explicit translation: 1 MiB = 1.04858 MB.

The full digital-storage comparison table lists every supported unit with its byte factor and a worked sample.

Pitfalls and gotchas

Decimal vs binary: 1 GB = 1,000,000,000 bytes; 1 GiB = 1,073,741,824 bytes. The two differ by 7.4%. Hard-drive labels use decimal; Windows file-manager labels use binary but call them 'GB'. The 1 TB drive that shows up as 931 GB free is exactly correct — both numbers, in their respective conventions, describe the same 1,000,000,000,000 bytes.

Bytes vs bits: 1 byte = 8 bits, since the 1960s. Storage capacity is quoted in bytes; network speeds are quoted in bits. A '100 Mbps' link transfers at most 12.5 MB per second of payload (and less after protocol overhead). Confusing the two leads to 8× speed-estimate errors.

kB vs KB vs KiB: lower-case 'k' is the SI prefix for 1,000; upper-case 'K' is sometimes used for the same value, sometimes for 1,024 — there is no consistent convention. IEC 80000-13 mandates lowercase 'k' for decimal kilobyte and 'Ki' for binary kibibyte. We respect that in our converter; in the wild, you have to read the context.

Storage vs memory conventions: hard drives and SSDs are quoted in decimal GB; RAM modules are quoted in binary GB. The same 'GB' can mean different things on two adjacent rows of the same spec sheet. When in doubt, multiply the rated GB by either 10⁹ or 1,073,741,824 and compare the byte counts.

Network advertising overhead: an ISP advertising '1 Gbps' is selling 1,000,000,000 bits per second of raw line rate. After Ethernet framing, TCP/IP headers, and retransmits, real-world throughput is typically 90–95% of that — so 1 Gbps caps real-world transfer at roughly 117 MB/s, not 125 MB/s.