A byte is the basic unit computers use to measure data, and on modern systems it is almost always eight bits. A bit is a single binary digit, either 0 or 1, and eight of them together can represent 256 different values, enough for a letter, a digit or a small number. File sizes, memory, storage capacity and data transfer allowances are counted in bytes, while network speeds are usually counted in bits. Mixing the two is one of the most common sources of confusion on IT quotes.
At a glance
- A bit is one binary digit (0 or 1); a byte is a group of bits, almost always eight.
- Lowercase b usually means bits and uppercase B means bytes, so 1 MB is eight times 1 Mb.
- Storage, memory and file sizes are counted in bytes; network bandwidth is usually counted in bits per second.
- Larger units step by 1,000: kilobytes, megabytes, gigabytes, terabytes and beyond. The binary units step by 1,024 and have their own names (KiB, MiB, GiB), though some software still labels them KB, MB and GB.
- In networking standards, “octet” is the precise term for exactly eight bits.
What problem it solves
Computers work in bits, but single bits are too small to be a practical unit for most work. Grouping bits into bytes gives a standard unit that can hold one character of text, and that memory, storage and file formats can be organized around. Most computers address memory one byte at a time, and file sizes, disk capacity and data transfer quotas are all expressed as byte counts.
For buyers, the byte is mostly a unit to read carefully. Storage arrays, backup licenses, cloud storage and data egress fees are priced per gigabyte or terabyte. Internet circuits and bandwidth are sold in megabits or gigabits per second. Comparing them, or estimating how long a transfer will take, means converting between bits and bytes.
How it works
Bits and bytes. A bit holds one of two values. Eight bits make a byte, which holds one of 256 values (2 to the power of 8). Four bits, half a byte, are sometimes called a nibble.
Why eight. Early computers used bytes of various sizes, including six and nine bits. The eight-bit byte became the norm with mainframe and then microprocessor designs from the 1960s and 1970s, and it is now effectively universal. Because the word “byte” was not always eight bits, networking and telecommunications standards often say “octet” when they mean exactly eight.
Notation. The common convention is B for bytes and b for bits. Providers and tools do not always follow it, so 100 MB on one screen and 100 Mb on another may or may not mean what the letters suggest. Check which unit is intended.
Larger units. Bytes are grouped into kilobytes, megabytes, gigabytes, terabytes, petabytes and beyond. Each step is either 1,000 times (decimal, the convention used by drive makers and networking) or 1,024 times (binary, used by memory and by some operating systems). The binary prefixes KiB, MiB and GiB were introduced so the binary sizes have their own names.
Converting speed to size. To estimate how much data a link moves, divide bits per second by eight. A 1 Gbps link carries at most 125 megabytes per second before protocol overhead.
When it matters for buyers
- Sizing storage. When comparing network-attached storage (NAS), SAN or cloud storage quotes, confirm whether capacity is raw or usable and whether it is stated in decimal or binary units. Our file and object storage page covers how buyers compare storage options.
- Planning data transfers. Migrations, backups and replication are sized in bytes but run over links rated in bits. Divide by eight, then allow for overhead.
- Reading usage bills. Cloud storage, file and object storage services and data transfer are billed per gigabyte, and the definition of a gigabyte should be in the pricing terms.
- Checking speed claims. A speed test showing 100 Mbps and a download showing about 12 MB/s are consistent; they are the same rate in different units.
Questions to ask vendors
- Is this capacity or usage figure in bits or bytes?
- Are your gigabytes and terabytes decimal (1,000-based) or binary (1,024-based)?
- Is the quoted storage raw capacity or usable capacity after formatting, redundancy and reserved space?
- For data transfer charges, how is a gigabyte defined and how is usage measured?
- What throughput in megabytes per second should we expect on this link for a large file transfer?
How it differs from a bit
A bit is the smallest unit of digital data, a single 0 or 1. A byte is a group of bits, almost always eight, and is the unit used for storage, memory and file sizes. Network speeds are normally quoted in bits per second, so a “100 Mbps” connection moves at most about 12.5 megabytes per second. When a number is written with a lowercase b, it usually means bits; an uppercase B usually means bytes.