Compare two type of addressing modes: $0-$address $($stack$)$ and $1-$address

$($accumulator$)$ architecture by writing programs to compute the Taylor series

expansion of cos$($m$)$:

cos$($m$)=1-$m$2/2+$ m$4/24$

Assume the following instructions are available:

$0-$address $1-$address

PUSH M LOAD M

POP M STORE M

ADD ADD M

SUB SUB M

MUL MUL M

DIV DIV M

DUP

Assume that you can replace any operand with a constant value, e$.$g$.$ PUSH $3,$ puts

the constant $3$ onto the stack. In these instructions, Opcodes are $8$ bits long; M is a

$16-$bit memory address; The data $($operand$)$ size is $32$ bits and constants are $16$ bits

long. The DUP instruction duplicates the top of the stack, so the same value appears

twice. Your comparison should be in terms of code size in terms of bytes needed.