Data Unit Converter

Convert between bits, bytes and every decimal and binary multiple, so the difference between a GB and a GiB stops costing you an argument.

Converter Bandwidth & Data Runs in your browser
Try:

Result

In bits
8,000,000,000,000
In bytes
1,000,000,000,000
Decimal (GB)
1,000 GB
Binary (GiB)
931.322575 GiB

Decimal prefixes (powers of 1000)

UnitValue
bit8,000,000,000,000
byte1,000,000,000,000
kbit8,000,000,000
kB1,000,000,000
Mbit8,000,000
MB1,000,000
Gbit8,000
GB1,000
Tbit8
TB1
Pbit0.008
PB0.001

Binary prefixes (powers of 1024)

UnitValue
Kibit7,812,500,000
KiB976,562,500
Mibit7,629,394.53125
MiB953,674.316406
Gibit7,450.580597
GiB931.322575
Tibit7.275958
TiB0.909495
Which prefix applies where
Network rates are always decimal: a 1 Gbit/s link is exactly 1,000,000,000 bit/s. Memory is always binary. Storage is decimal on the box and often binary in the operating system - which is why a "2 TB" disk shows as 1.82 TiB.

About Data Unit Converter

Two prefix systems, one letter apart, differing by up to 10% at terabyte scale. This converts between all of them explicitly, so a capacity plan and a vendor datasheet can be compared without guessing which convention each used.

Converting a data cap into a sustained rate

Capacity questions usually arrive as a volume and need answering as a rate. One terabyte per 30-day month is 8 Tbit spread over 2.592 million seconds, about 3.09 Mbit/s sustained — which is why a 1 TB cap and a gigabit line are not in tension for typical use. Run the same arithmetic before sizing a backup window: 4 TB overnight in eight hours needs 1.1 Gbit/s of goodput, so a gigabit link cannot do it regardless of tuning.

Decimal or binary?

IEC binary prefixes - KiB, MiB, GiB, TiB - are powers of 1024. SI decimal prefixes - kB, MB, GB, TB - are powers of 1000. The gap compounds: 2.4% at kilo, 4.9% at mega, 7.4% at giga, 10% at tera. Networking uses decimal throughout, memory uses binary, and storage marketing uses decimal while Windows reports binary under decimal labels.

Common use cases

  • Reconciling a disk vendor capacity with what the OS reports.
  • Converting a monthly data cap into a sustained bit rate.
  • Checking whether a monitoring graph is showing bits or bytes.

Edge cases and gotchas

  • Lowercase b is bits, uppercase B is bytes - an 8x difference hiding in one character.
  • Windows displays GiB but labels it GB; Linux and macOS are generally explicit.

Frequently asked questions

Why do networking and storage use different prefixes?
Networking has always counted in powers of ten because signalling rates are decimal. Memory addressing is inherently binary. Storage sits between the two, marketed in decimal and often reported by operating systems in binary.
Why is my 2 TB disk only 1.8 TB?
It is 2 TB decimal, which is 1.82 TiB binary. Windows shows the binary value with a decimal label. No capacity is missing.