Computer systems today primarily store “words” of data in either Little or Big Endian format, in the physical memory. By “word” I mean a piece of data that is larger than a byte. Typically a word is defined as 2 or 4 bytes, and halfwords, double words, quadwords are extrapolations thereof. Intel machines store data in their memory, in Little Endian format and IBM PowerPC architecture stores data in Big Endian format. The differences are critical when writing a compiler or when transferring a piece of data written for one system to another.
Simple way to remember is: Little Endian means, the little end (least-significant byte) of the multi-byte word goes into memory first. First here means a lower memory address. And Big Endian is vice-versa.
Infact, the origin of these words is from Jonathan Swift’s Gulliver’s Travels, where in the Lilliput kingdom, a decree by the King, that eggs be broken from the little side, caused a rebellion by a faction called the Big Endians, who were used to breaking their eggs from the bigger end.