The aam (ASCII Adjust after Multiplication) instruction, like aaa and aas, lets you adjust an unpacked decimal value after multiplication. This instruction operates directly on the ax register. It assumes that you've multiplied two eight bit values in the range 0..9 together and the result is sitting in ax (actually, the result will be sitting in al since 9*9 is 81, the largest possible value; ah must contain zero). This instruction divides ax by 10 and leaves the quotient in ah and the remainder in al:
ah := ax div 10 al := ax mod 10
Unlike the other decimal/ASCII adjust instructions, assembly language programs regularly use aam since conversion between number bases uses this algorithm.
Note: the aam instruction consists of a two byte opcode, the second byte of which is the immediate constant 10. Assembly language programmers have discovered that if you substitute another immediate value for this constant, you can change the divisor in the above algorithm. This, however, is an undocumented feature. It works in all varieties of the processor Intel has produced to date, but there is no guarantee that Intel will support this in future processors. Of course, the 80286 and later processors let you multiply by a constant, so this trick is hardly necessary on modern systems.