littleendian

Little-endian refers to a convention for ordering bytes within multi-byte data types in a computer system. In a little-endian system, the least significant byte of a value is stored at the lowest memory address, with more significant bytes located at higher addresses. For example, the 32-bit value 0x01020304 would be stored in memory as 04 03 02 01. The interpretation of the same byte sequence depends on the endianness of the system that reads it.

Most modern CPUs are little-endian, including x86 and x86-64 processors, as well as many ARM implementations. Some architectures support big-endian ordering or operate in either mode (bi-endian or mixed-endian), such as SPARC and certain PowerPC configurations. ARM can run in either little- or big-endian mode, selected at system or process startup. Endianness also affects data serialization, file formats, and inter-system communication, making cross-platform data exchange a common source of bugs if not handled explicitly.

The terms little-endian and big-endian originate from a naming convention popularized in the late 20th century, drawing on a Gulliver’s Travels reference to different factions preferring different ends of an egg. The exact historical origin is debated, but the labels have become standard descriptors for byte-order differences across architectures.

In networking, the standard is big-endian, often called network byte order. Byte-order conversion functions (for example, htonl/ntohl in C, or similar utilities in other languages) are commonly used to ensure consistent interpretation when data is transmitted between heterogeneous systems.