Buy a "1 TB" external drive, plug it in, and the operating system reports something like 931 GB. Browse a download page advertised as "100 Mbps" and the file actually arrives at around 12 MB per second. Neither of these is a bug; they are both consequences of two related conventions that share the same letters and mean different things. This guide takes the conventions apart so the numbers stop being mysterious.

Bits and Bytes

A bit is a single binary digit — a zero or a one. A byte is a group of eight bits, large enough to hold one character of basic text or a small integer. Almost all stored data is measured in bytes, while almost all data in transit is measured in bits per second. That is the first and most common source of confusion: a "100 Mbps" connection is 100 million bits per second, which is roughly 12.5 million bytes per second, before any protocol overhead.

The standard symbols are lowercase b for bit and uppercase B for byte. Mixing them up changes a number by a factor of eight. 50 MB/s and 50 Mb/s are not the same speed.

Decimal Prefixes (SI)

The everyday metric prefixes — kilo, mega, giga, tera — are powers of 1000. They are part of the SI system and are used universally for things like distance, mass and frequency. When applied to storage:

• 1 kilobyte (kB) = 1,000 bytes
• 1 megabyte (MB) = 1,000,000 bytes
• 1 gigabyte (GB) = 1,000,000,000 bytes
• 1 terabyte (TB) = 1,000,000,000,000 bytes

Hard-disk and SSD manufacturers use these decimal definitions on the box, and so do most network speed quotes.

Binary Prefixes (IEC)

Computers, however, naturally count in powers of two. Memory chips, file systems and operating systems use the binary forms:

• 1 kibibyte (KiB) = 1,024 bytes (2¹⁰)
• 1 mebibyte (MiB) = 1,048,576 bytes (2²⁰)
• 1 gibibyte (GiB) = 1,073,741,824 bytes (2³⁰)
• 1 tebibyte (TiB) = 1,099,511,627,776 bytes (2⁴⁰)

These were standardised by the International Electrotechnical Commission (IEC) in 1998. They sit alongside the SI prefixes and are designed to remove ambiguity. The names — kibi, mebi, gibi, tebi — are deliberately a little awkward so they cannot be confused with kilo, mega, giga, tera.

Why a 1 TB Drive Shows as 931 GB

The drive really does hold one trillion bytes — 1,000,000,000,000. That matches the manufacturer's "1 TB" label exactly, using the decimal SI definition. Windows reports drive capacity using binary gibibytes but still labels the unit "GB". So the same one trillion bytes becomes:

1,000,000,000,000 ÷ 1,073,741,824 ≈ 931.32.

The drive has not lost capacity. The two numbers describe the same amount of storage in different units that happen to share an abbreviation. macOS and most Linux file managers now use the SI definitions and would simply show "1 TB" — a more honest label, even if it sometimes confuses users moving between operating systems.

Network Speed: Where the Confusion Compounds

When an internet plan is advertised as "100 Mbps", the unit is megabits per second, not megabytes per second. A back-of-envelope rule:

• Divide bits per second by 8 to get bytes per second.
• Round down a little to allow for protocol overhead (TCP/IP headers, retransmission, encryption).

A 100 Mbps line will deliver a file at perhaps 11–12 MB/s on a sustained download. A "1 Gbps" fibre line tops out around 110–120 MB/s in practice. If a download is much slower than that, the bottleneck is somewhere else — disk write speed, the source server, or wireless link quality.

A Quick Conversion Table

• 1 byte (B) = 8 bits (b)
• 1 KB (decimal) = 1,000 B
• 1 KiB (binary) = 1,024 B
• 1 MB = 1,000 KB = 1,000,000 B
• 1 MiB = 1,024 KiB = 1,048,576 B
• 1 GB = 10⁹ B; 1 GiB = 2³⁰ B (about 7.4% larger than 1 GB)
• 1 TB = 10¹² B; 1 TiB = 2⁴⁰ B (about 10% larger than 1 TB)

The percentage gap between the decimal and binary definitions widens at every step: about 2.4% at the kilo level, 5% at mega, 7% at giga, 10% at tera. That is why the "missing capacity" looks bigger on big drives.

Common Mistakes

• Mixing bits and bytes. Always check whether a unit is b or B. The factor of eight is huge.
• Comparing two systems' capacity reports. A drive may "shrink" or "grow" depending on which OS you plug it into; the bytes are the same.
• Forgetting overhead. File-system metadata, redundancy and reserved space all consume some of the total. A 256 GB SSD does not give you 256 GB of usable space.
• Using "K" or "M" without context. In RAM specifications, "K" historically meant 1024. In file sizes, modern operating systems are inconsistent. When in doubt, treat the symbol as ambiguous and check.

Worked Example: Sizing a Backup

You have 480 GiB of data on your laptop and want to know whether it fits on a "500 GB" external drive.

• 480 GiB × 1,073,741,824 ≈ 5.153 × 10¹¹ bytes — your data.
• 500 GB = 5 × 10¹¹ bytes — the drive.
• Your data is about 3% larger than the drive's capacity. The drive will fill before the copy finishes.

To be safe, pick a drive whose decimal capacity is comfortably above your binary usage — a 1 TB or 2 TB unit, in this case.

Where to Go From Here

For specific conversions, the ConvertMe.org home page has dedicated digital-storage and data-transfer categories. For the broader prefix system, see the SI base units and prefixes guide. The metric vs. imperial guide covers the wider question of why the same word can mean different sizes in different traditions.