In what context?
Where have you seen the term used? - is this (part of) a homework assignment?

In the context of 'C' (and probably other) compilers, padding is added by the compiler when the target processor has specific requirements about how multi-byte data objects are aligned in memory.

eg, processors with word sizes of 16 bits or more often require that 2-byte objects start on an even address; or, if they don't require it, they perform very much better when it is so.

similarly, 4-byte objects may be required (or preferred) to start at an address that is a multiple of 4.

etc, etc.

There is usually a control to override this.

None of this matters to an 8051, because it can only ever access 1 byte at a time - so it makes no difference what address (odd, even, etc) it starts on!

As Neil noted, the only time it may affect a system with an 8051 is when the 8051 part needs to share data with an alignment-sensitive part...