Endianness: Big-Endian vs Little-Endian Byte Order

Endianness describes the order in which bytes of a multi-byte value are stored in memory. This seemingly simple concept is a major source of bugs in cross-platform software, network programming, and binary file parsing.

Big-endian (network byte order):

The most significant byte is stored at the lowest memory address. The number 0x12345678 is stored as:

Address:  0x00  0x01  0x02  0x03
Value:    0x12  0x34  0x56  0x78

This matches how humans write numbers (most significant digit first). TCP/IP network protocols use big-endian, which is why it is also called "network byte order."

Little-endian (Intel byte order):

The least significant byte is stored at the lowest memory address. The same 0x12345678 is stored as:

Address:  0x00  0x01  0x02  0x03
Value:    0x78  0x56  0x34  0x12

Most modern CPUs (x86, x86-64, most ARM configurations) use little-endian.

Why little-endian exists:

Little-endian simplifies certain hardware operations. When you cast a 32-bit integer to a 16-bit integer, the value is already at the same memory address — no pointer adjustment needed. Addition also proceeds naturally from the lowest byte upward.

When endianness causes problems:

• Network communication: Data sent from an Intel machine (little-endian) to a SPARC machine (big-endian) will be garbled unless byte-swapped. Functions like htonl() and ntohl() handle this.
• Binary file formats: File formats specify their endianness. BMP files use little-endian. TIFF files include an endianness marker. Reading a file with the wrong byte order produces nonsensical values.
• Hexdump analysis: When reading a hexdump of memory from an x86 system, a 32-bit integer 0x00000001 appears as 01 00 00 00, not 00 00 00 01.

Detecting endianness in code:

In JavaScript: new Uint8Array(new Uint32Array([1]).buffer)[0] === 1 returns true on little-endian systems. In Python: import sys; sys.byteorder returns 'little' or 'big'.